Hey, good morning. Welcome to Stem 2022 Investor and Analyst Day. I'm Ted Durbin. I'm the Head of Investor Relations at Stem. We're thrilled to host you at this inaugural event for us. Welcome to everybody here in the room in New York, and to those of you listening on the webcast, thank you for taking the time to be with us. Before we get started, I'm gonna say please silence your cell phones and laptops. Thank you. A couple housekeeping items. The presentation we're gonna use today will be posted to the investor relations portion of our website this morning. It's at www.investors.stem.com. We have about 2 to 2.5 hours prepared remarks, and that'll be followed by a question and answer session.
For those of you in the room, we do ask that you hold your questions until the end of the presentation, and we're gonna be compiling questions on the webcast as well, and so we'll try to get to as many questions as we can at the end of the session. Lastly, before we begin, I would like to remind all participants that some of the statements we'll be making today are forward-looking. These matters involve risks and uncertainties that could cause our results to differ materially from those projected in these statements. I therefore refer you to our latest 10K filing and our other SEC filings. Our comments today will also include non-GAAP financial measures. Additional details and notes on the reconciliation of these non-GAAP measures to the most directly comparable GAAP financial measures can be found in the appendix of the presentation.
With that, I'm excited to turn the floor over to John Carrington, our Chief Executive Officer. Thank you.
Thanks, Ted. I'll second your welcome. I really wanna thank everybody for joining, particularly the shareholders, analysts. We have a good Stem contingent that you'll hear from later today, and I know there's a significant crowd on the webcast, so we appreciate you taking the time to spend the next few hours with us. We are really, really excited about the agenda that we've put forth. I couldn't be more proud of this team. I think you're going to get a real sense today of how broad and the depth and breadth of this team. You're gonna hear a lot about the Athena differentiation. You could see the agenda here. I'm gonna do a strategic overview for about 25, 30 minutes, and then we'll jump into some of the growth outlooks with Prakesh.
He'll talk more about the IRA and the implications there and some strategies we have there. Also, a nice opportunity to meet Bob Schaefer. He's the co-founder of AlsoEnergy, which was the acquisition that we completed earlier in the year. I think many of you are aware of that, but it'll be nice to have a chance to hear from Bob. On the technology side, Larsh will have a very extensive overview, as I said, on Athena and the roadmap and path in which we're going forward. Then we'll have a break. We'll come back, and then Bryan Ho is going to walk through some software demos, and I think you'll find that very instructive. Show a little bit of the Athena portal and some of the value that really brings to our customers and how customers use the product.
Finally, we'll have a technology roundtable on the technology front, and I think you'll find that very interesting, whereby we bring the product and the software developers together to really, again, help you understand how we go to market, how we think about building our software and the timelines we think about, and kinda the lean philosophy that we use. Then Bill will wrap with a finance update. Then, as Ted mentioned, we'll go right into Q&A. Let me jump right into my slides. I'm gonna talk a little bit about the team itself. I'll go through the 4 ways in which we think about how we will win as a company, and then also a final takeaway slide on some of my thoughts going forward.
From a leadership team, look, I think it's one of the best teams in the industry. The individuals across the top I just mentioned are the presenters today. The balance of the individuals on this page represent the executive team. I would highlight that Mike Carlson is here in the back left. At lunch, maybe you have a chance to say hello to him. He's our new Chief Operating Officer that we've announced recently. Excited to have Mike. Comes to us from Coke, has also spent time at GE, ABB, and Siemens, so a terrific addition to the team. I would say this is a team that's worked together for many years for the most part. We have also added some incredible new people, all of which, though, have extensive experience in software, industrials, and in the energy space.
I think a really top-notch team with some great experience to help us go into the next level of this business, welcoming them. You know, I think one of the big components of how we think about the company is around execution, and this is the core guiding principles that we use. It's really off of OKR, so you think about objectives and key results. Each one of these is an objective. Each one of these is owned by 1 or 2 of the executive staff members that you saw on the previous slide. What we do is we build key results to the entire business around each of these objectives. An individual in India could look at this and understand what result they're responsible for to execute on that objective.
I think it's very important that everyone's aligned to our success, and they understand what they have to do to enable these objectives. These are reviewed on a weekly basis. It's a rigorous process, and again, it's really focused on our continued commitment to executing our goals. I think a lot of you haven't been with the company through the journey to get to today. I know some. I saw one of our bankers that's been on that journey smiling as I said that. Look, I think it's important to kinda recap a little bit about what got us here. It's a history of execution. It's a somewhat remarkable path that we got. It was never a straight line as it typically isn't.
You know, when we were in California on the C&I market, we were really the first to introduce an ESS product. It was so early, in fact, that we had to build the inverters. We put batteries into a server. If you think of a server box, put 18 batteries in and very manual process. That box saved anywhere from $7,500 to $10,000 a year. Okay? We had customers say, "It costs me more to do the contracting than put this product in." Those were the early days of this business, but it was always grounded in software, and it was also grounded in building out virtual power plants. We started to execute on the Fortune 500s. We took it to aggregating all of these and participating in the wholesale market. We participated in day-ahead and real-time markets.
We were the first to do that with a battery system. We eventually developed the largest virtual power plant in California with a utility, and that's with Southern California Edison. You can see 420 MWh. I think it was a very admirable feat that we accomplished in California. Then we started to get a lot of incoming in front of the meter, and people said, you know, "Do you think you can really execute on front of the meter because you've done such a great job on behind the meter?" Well, we went from 0% to over 50% market share in 18 months in the ISO New England. A real accomplishment that is very much attributable to our software platform because we could translate our behind-the-meter software deck to that front-of-the-meter need.
I think that was a big milestone for the business to make that transition. We also acquired the number 1 solar asset performance management software platform. You'll hear more about this from Bob. This is the AlsoEnergy piece, and it's just a really great addition to the Stem family, and you'll see some of the upside associated with that. Then finally, we announced last quarter that we're the only software provider in the space to bring all 13 Rocky Mountain Institute value streams as a software offering. I think it's good to get a little level set on some of the key milestones that got us here. On the right side, you can see how our execution has really generated significant enterprise value over the last several years.
The right side might be my favorite slide I've ever seen or done, so I'm pretty excited about that one. Again, we expect a continued up and to the right as we go forward. Along the lines of execution, look, as I said, we have really built a foundation for success, and I think it really shows in the results. As you can see here, revenues have nearly tripled over the last 3 years. Bookings on track to nearly double. These numbers are, I think, exceptional. I think what's really exciting about it is all of this drives strong recurring long-term software revenue. The difference between the $187 and the $169 as an example on the revenue side is just low end and high end of guidance on a CAGR level.
Just to level set the footnotes are probably difficult to see here in the audience. If I go over to kind of the money slide, you know, today we are announcing that the company will be EBITDA positive in the H2 of 2023. We are very focused on our growth in software and services, and we believe this drives 15% to 20% long-term EBITDA margin. Thirdly, we are reaffirming our 2022 guidance. Going forward, 4 ways in which we will win. Technology leadership, operational excellence, commercial strategy, and micro tailwinds. Macro, excuse me, macro tailwinds. From a technology leadership standpoint, this is an analysis that we've not shared before. It's 1e that McKinsey did.
They have a technology team specifically aligned to analyzing your digital and analytical capabilities, and it's a deep group of subject matter expertise in the software space. What they did is looked at us, deep dive on the Athena platform, and then they looked at how we scored versus 550 other companies. We scored higher than the average digital leader. Now you may say, who's the digital leader? Well, the digital leaders included Google, Amazon, and Netflix. These are your top software developers and digital leaders globally. As you can see where Stem landed on that front. From the perspective of some of the items that they measured us on the core dimensions, it was really an extensive analysis done.
You could see strategy, org, and talent, agile delivery, which you'll hear more about from Albert later today, capabilities, adaptation, adoption, and scaling. I would say Stem's technology team is and the group is very highly performing. You could see that we're on the high end across every 1 of these 5 metrics and, you know, you'll see the tangible results of this later today as we go through the software and the Athena platform in more detail. From a, you know, combined company standpoint, I think we are very uniquely positioned to build a platform-based model. The exciting part is our customers want a single pane of glass, and we're gonna bring that to them. There's no other software provider that has the solar and storage optimization capabilities, longevity or amount of assets under management that Stem brings now.
This is a very exciting acquisition that we've executed on. I would say also, not only were they nominated number 1 by Guidehouse again, but STEM was named 1 of the top 10 most innovative energy firms in the world by Fast Company. I think we're both getting a lot of exposure and being recognized for the software platform and the execution that we've accomplished. From an operational excellence standpoint, you know, I would say our technology teams have really accomplished some notable milestones, including 13, all 13 of RMI, as I mentioned, value streams on the RMI wheel. From a, you know, utility obligation standpoint, look, you see it today in Florida, but we've seen many market issues, catastrophic weather conditions, and strain, and these are straining the grids.
Our utility customers are calling on our systems more than ever. We are becoming a more and more material part of helping them in times of need, and we're executing on 100% of our utility obligations, and we're getting called on year-over-year over 70% more. It's just becoming a remarkable, important component of the grid. On the ISO New England side, we've built a leading market share, as I mentioned, by delivering better forecasted financial performance, achieving over 96% of perfect forecast revenues. That's a very impressive number. It's best in class. Then from a greenhouse gas optimization standpoint, this is a very interesting newer component of Athena.
What we've done is we've provided our customers a software solution to look at all the types of energy, sources of their energy, and we can actually utilize Athena at the right time to help drive their greenhouse gas emissions lower. Our customers are coming to this for this service. They wanna understand where the energy is coming from, what type of generation, and utilize Stem system to offset that greenhouse gas issue. It's something that we're working on with our C&I customers and having a tremendous amount of success. Then finally, you know, the solar asset performance business is very differentiated. You'll see here as well as you'll hear more from Bob, but we continue to be the market leader in that space with the PowerTrack technology.
On the booking side, I would say that, you know, the continued momentum around bookings, demand continues to be strong. We'll talk more about that with Bill later in the discussion. You know, as we build our pipeline of software-only deals and really flex our power pricing, you'll see some price increases that we've done, and we've talked about this on earnings calls. You know, we continue to deliver growing gross margins with accelerating software share, and we see that as driving higher gross margins throughout the model period. We also see strong growth in our contracted annual revenue, annual recurring revenue, our CARR metric that we introduced some time ago. Then finally, on the operating leverage standpoint, this is an area Bill will talk more about.
I'm really excited about this because we have over 100 people in India that came with the AlsoEnergy acquisition. Having that installed base and that center of excellence that we're going to build there and expand there will really help us drive down the OpEx piece of our business. As you can see, we've went from 44% in 2021 as a percent of revenue, and we're at 33% as the midpoint of guidance for this year. On the commercial side, I mean, I think a lot of you have seen this page. You know, 3 pillars to our model. Software drives 80% gross margin with long-term contracted recurring revenue. The hardware continues to be a strategic subject matter expertise product that we bring to our customers. Notably, it's 1 that we can serve in a very low-cost way.
In fact, we only have about 10 people focused on hardware in this company, and 100% of our hardware sales are software attached with Athena. It's an important part of our business, and again, we view it as a low cost to serve. Again, on the market participation, we see long-term upside from the aggregated capacity that we have in the market, and this continues to drive higher than forecasted project economics for our customers. Market participation is an important part of our model as we go forward. From a value prop standpoint, there are really 2 segments of the market that we focus on, behind the meter and in front of the meter, and we're continuing to build out our EV offering. We'll talk a little bit about that as well.
That's an important component, particularly for a lot of the behind-the-meter customers that you see here and others. We really built the behind-the-meter market. Today, I think many of these Fortune 500 customers think of us as a trusted advisor. They think of us as someone that can help them install more storage, help them with the solar plus storage, also help them think through their EV and fleet electrification. A lot of our customers are trying to understand that piece in companies like UPS and Home Depot in particular. We'll continue to be that trusted advisor to them and we feel like we have the right offering to execute in that key space for us.
Finally, on the front-of-the-meter side, this is a market, as I mentioned, that we really continue to have great success on. I'd say that, you know, we have simplified the solution to help these customers achieve significant increases in their returns. You see the IRR increases of 10% to 40%. We have the software platform that they need. We'll continue to grow this. We are becoming the de facto company for the space. As I mentioned, you know, really to go from 0 to leading market share in 18 months in ISO New England is very impressive. It says a lot about the software. It also says a lot about Athena that over 50% of our contracted volume last quarter was from repeat customers. We are executing.
The say-do ratio is there, and they're continuing to come back and purchase from Stem. From a macro tailwind standpoint, look, I think everyone's very familiar with Inflation Reduction Act. This is significant for the entire industry, but particularly for storage. We feel like we're incredibly well-positioned in this area. We're gonna continue to work through the details. You'll hear that today from Bill. I mean, we still have to unpack some of the tax components around IRA, but all in all, it's a nice opportunity for the business, certainly. There's 3 areas I'd highlight. On the greenfield side, we see customer economics improving from 30% to 60%. One of the unique things about the AlsoEnergy acquisition is that we have over 40,000 existing customers that are solar only.
This is a very compelling competitive moat for our sales team to go see those Also customers, add storage to those existing solar sites. We know where they are, we know how big the solar system is, we know how to size the system. As you can see here, we view it as a $6 billion potential in that portfolio of customers. No other competitor has that type of an opportunity and just a hunting ground, if you will, from a sales standpoint. Then finally, on the solar production tax credit, we believe that, you know, with Athena and PowerTrack, we enable over 500 basis points. Prakesh will talk through that more in his section. He unpacks this.
It's a very interesting offering that we think we can bring our customers, and we're gonna share that with you today. On the battery supply chain piece. Look, I think, you know, IRA will drive further U.S. manufacturing. We're seeing the announcements already. I expect significant capacity expansions globally in addition to just the U.S. You know, from a Stem standpoint, we're fully contracted through 3Q 2023. We're executing on the Q 4 of 2023 and starting to look at the beginning of 2024. As we've told you, we typically try to be a year out, and unless there's some big dislocation that would allow us to go longer or we change that strategy. Typically, that's how you can think about our perspective on our supply chain and how we kinda contract that out longer term.
I'd say that look, despite near-term fluctuations in supply chain, the long-term trajectory for battery prices is deflationary, as you can see here. 20% price down, that goes out to about 2030. Big cost component is still the cells, so we continue to monitor that very closely. I think that, you know, we've added some very strong supply chain team members, and I think that will really help continue the momentum that we've had in this space. I would say another thing to note is, in the last couple months, I think there's been over a dozen gigafactories announced in the U.S. alone.
The amount of capacity coming on, if you were one of the 28,000 people that participated in RE+ last week, you saw and heard how many suppliers are adding capacity, I'm sure, and it's an impressive amount. Now, I don't think the U.S. piece will be on for a couple of years. That's okay, because I think everyone globally is gonna continue to add capacity. 3 takeaways here that I just wanna hit on before I turn it over to Prakesh. Number 1, this really is a massive market. You know, energy storage and asset performance management are some of the fastest-growing areas, and we really offer the potential for the most profitable market opportunity.
Really, we view it as kind of the broadest energy landscape, where we can bring solar and storage together and execute on that for our customers. You're gonna hear a lot about the innovative products and services today, and I think you'll walk away realizing how differentiated we are. We really believe that both the team and the platform is unique. You saw it in the McKinsey report. You're gonna see it here live today, and we're really excited to share all that with you. Then on the profitability standpoint, you know, we really view this as a core competency, and you're gonna see, you know, our commitment to financial discipline and really, you know, we expect to meet these profitability goals that we've talked about in the H2 of next year.
There's gonna be significant financial discipline. Bill Bush and his team are well-staffed to execute on this. You know, we feel like driving a capital-light business that can generate a significant free cash flow is very important for our investors and the company in general. With that, I'll turn the discussion over to our Chief Strategy Officer, Prakash Patel.
Great to be with you today. Gonna share a couple of our thoughts on our growth outlook and, in particular, a preliminary view on how we expect the Inflation Reduction Act to impact our business and our view on growth. We'll provide some upside opportunities as John mentioned. We think there's some really compelling new offerings that we can bring to market, and we'll detail some of the returns and how we think that's going to hit with customers. You know, wanted to lay out. This is really a preliminary view. As John discussed, there's still a couple of considerations around the IRA that need to get resolved by the market. Overall, we think the impact of the Inflation Reduction Act is really underappreciated by the market.
As you'll see in Bill's slides, there are several factors that we think once resolved could cause us to have a more aggressive view on our growth outlook. Based on current estimates on just the impact of the standalone storage ITC, you know, based on market estimates that we've seen, we expect about a 50% or greater increase in the rate of growth across both the behind the meter and in front of the meter markets. Then we'll unpack the expected growth in the solar asset performance market as well. Overall, you know, continued strong growth and a lot of potential tailwind that could accelerate that further.
Here, one of the points I wanted to highlight is really the fact that, you know, we haven't seen a complete view on the impact of the Inflation Reduction Act come out from third-party research houses. WoodMac released something just around the standalone storage ITC, but there are several provisions around the broader market that we think could be synergistic in driving demand. I'll list some of those here. In particular, the extension of the solar tax credits, wind, and many of the provisions around electric vehicles. Each of these have a synergistic effect on driving more demand for energy storage, and we'll talk about how we're capitalizing on these opportunities. Overall, we think we're very well-positioned to bring to market these solutions, and we'll talk about some of that in a couple slides.
The first segment, you know, really where Stem started, the behind the meter market. We expect the benefits of the IRA to have a particularly meaningful impact in expanding our business to drive greater volume in this segment. As we've discussed, we have significant, you know, deep relationships with many of the Fortune 500s. Recently we've had a lot of incoming requests from these accounts around thinking through national rollouts across their footprint. You know, historically, we were primarily focused in markets where storage penciled, such as California, New York, and other geographies. Today, we're in discussions with many brands around a national rollout.
You know, I think we've mentioned with some of the folks here, if you think of just one of our accounts as an example, Home Depot, their footprint across the U.S. represents a several hundred million dollar opportunity alone. Now, when you multiply this across the, you know, several dozen accounts that we have in the BTM market, we think that's a very compelling number, and we're very optimistic about our prospects for capturing growth in this segment. One of the aspects that we think could be, you know, really interesting is the electrification of electric vehicle fleets.
On the right-hand side, this is an analysis from NREL, where basically the takeaway is that the expected load growth from the electric vehicle fleet market is expected to be as large or larger than the existing commercial and industrial load in the US. Effectively, this means that the BTM market should double in size as far as an opportunity for us for helping these customers to reduce their electricity bills, and in particular, drive sustainability solutions across Fortune 500s and the broader BTM market. Our view is conservatively about 50% of our BTM business will be attached with EVs. Notably, as we talked about, you know, the system sizes should be much larger in these opportunities by 2x to 5x.
These installations also come with compelling hardware gross margins and about a double of our typical software fee on installations at these locations. Next, wanna talk about the front of the meter market. This is something as John mentioned. Put this over here. As John mentioned, you know, we entered recently about 3 years ago, and we continue to see strong momentum. It's the fastest-growing segment of the storage landscape. 58% compound annual growth rate forecast by Wood Mackenzie through 2025. This represents about 900 basis points better than what we showed in our prior analysis day about a year ago. Notably, it's almost a double of what was forecast for the 2025 period. A lot of acceleration in this market. You know what?
Across both the front of the meter and behind the meter space, one of the things we're really excited about is the fact that market participation revenues, we're expecting that to be significantly a greater share of our project economics. That's really because we no longer have this restriction on our trading activity. Historically, we were relying on the storage ITC. Now with the standalone storage ITC, there's no longer that restriction on charging. As a result, we expect greater market participation revenues in systems across BTM and FTM. Lastly, these are forecasts on the solar asset performance monitoring segment, both in the U.S. and globally.
I'll remind everyone of the bulk of AlsoEnergy's businesses in the U.S., about 90%, and with the majority of that in the behind the meter segment. We expect a lot of growth there. You know, John talked a little bit about the opportunity with their captive footprint, and I'll go into detail on that in the next slide here. You know, one of the things we did post closing the acquisition is look through the install register at AlsoEnergy, and our sales teams have begun outreach on discussions around retrofitting existing solar locations with storage. With the storage ITC, this is very compelling economics.
If you look at just the core markets where Stem is currently active, that represents about a $6 billion opportunity across the entire AlsoEnergy footprint. That's 20 billion-dollar customers where we have their solar data. We have a contractual relationship with them. They're very happy with our service, and it's very easy for us to access that market. We're excited, in particular, to think of ways how we can efficiently target this market, right? This is 41,000 sites, and so it really takes someone who's serviced a fragmented set of customers that Stem has through its history. We're focused on automated tools and customer self-service offerings to programmatically sell into this install base.
If you think about the portfolios of other asset managers, you know, like a GSAM or Brookfield, we think we can replicate this programmatic selling capability across their footprint as well, and leverage that same strategy across the Fortune 500. At a minimum, just within our customer base, it's several billion, and we think we can multiply that across the broader C&I landscape. This slide. You know, this is really something that we think is uniquely differentiated that we can bring into the market. On the right-hand side, this is something that several market analysts have put together. Essentially, what it is is the.
What it represents is one of the provisions of the Inflation Reduction Act that relate to the ability to opt into the production tax credit in lieu of the solar ITC. What's required to receive this incentive is really a software solution that can ensure credible uptime and generation across the asset life, whereby a developer would take the incentive over time rather than the upfront ITC incentive. What this analysis says, and it's really you know difficult to decipher in this table, but what we've done separately across our existing customer base is to use the pro forma project economics that we have and looked at a scenario where we took the solar PTC.
The key conclusion is that we can enhance project returns by anywhere from 500 to 1,000 basis points on these systems by electing to take PTC. That's a uniquely differentiated financial outcome that's enabled by the combination of Stem and AlsoEnergy. Again, something that we can uniquely bring to market, substantially better returns in select instances. This bundle offering is one we're taking to market right now and really capitalizing on some of the provisions of the Inflation Reduction Act. Lastly, as I've been mentioning, you know, we still and industry analysts are still digesting some of the provisions of the Inflation Reduction Act. Bill will talk about some of the factors that lead us to guide around the growth rates that we're expecting.
However, we expect to revisit that as some of these items, including the battery supply chain, as an example, resolve. We're pretty optimistic. As John mentioned and have on this page, you know, there's a 13x expansion in domestic manufacturing capacity that's been announced. About 10 gigafactories have been announced as well in the U.S. There's several factors that we think should drive substantial demand for storage. We're really excited. Next, I wanna hand it over to Bob Schaefer and then Larsh Johnson to talk about exactly how our technology solutions will capitalize on this opportunity. I'll hand it off to Bob.
Thanks, Prakesh. Good morning, folks. My name is Bob Schaefer. I'm president of AlsoEnergy, one of its founders. I'm real excited to be here. I'm excited to share with you our vision, and really the joint vision that we have here together. You know, when I would start my meetings with the team, we talked about our vision of the company and our vision of the future. Honestly, that vision was a vision where all generation on the grid was renewable, driven by cost and by resilience. We're gonna see kinda the story and how it evolves to where it is today, where we can achieve that vision. First, you may be wondering what AlsoEnergy is and what we do. What we do is we provide a software platform called an Asset Performance Management Platform.
What that does is it gathers data from sites, in AlsoEnergy's case, over 200,000 sites, located wherever they are in the world, in 50 countries, in fact, gathers that data and allows people to monitor, manage, control, and increase the value of those properties, right? Those assets, those investments. Without the products and services that we provide, you'd manage things with paper and pencil, spreadsheets, you know, maybe some other Microsoft products. It would be awful. Absolutely awful. You know, really what we developed here is we built a system of record, right? It's a system of record where all primary data associated with a site performance is located. That allows us to do analytics, allows us to do optimization and reduce the cost to maintain, while at the same time increasing the performance, increasing the financial performance of a system.
You know, prior to AlsoEnergy, I spent 25 years in the computer industry. We just learned about best practices, right? We were in a fast-moving industry. We had best practices. We figured out best practices for products, for teams, for companies, and how to grow them. What we did when we started the company in 2007, we thought we had the expertise to make a difference. That difference really was we would apply those best practices to an emerging industry, right? The emerging industry being renewables. It took us a little time to figure it out. We started in 2007. Here we are today. I think we've done a great job. The company is successful and our flagship product, PowerTrack, is a system of record for solar PV sites.
Here you might wanna wonder about, like, how we got here, right? What happened? You know, honestly, you know, solar is still a maturing industry, but we've been around since the beginning here in the United States. We've grown with the industry. We've remained trusted and a reliable name. We've been there through the ups and the downs, and we'll be here tomorrow for you. We're a stable company, and that's important because the assets that our customers are deploying have an expected life of 20 to 25 years. It's really kinda critical. The recent acquisition of our company by Stem really provides us further stability and really a further growth opportunity. You know, since we started and had our first revenue in 2009, 2 years after we started the company, we've grown to become a global provider.
We started with an acquisition in 2013, which was done on the balance sheet, which really helped us kind of understand and grow in the C&I market here domestically. You know, when I talk about markets here, we talk about 3 basic markets, a residential market, homes, systems on rooftops. We talk about a C&I market, which is typically systems under 5 MW in size. We talk about utility markets, which are those markets and those sites over 5 MW. When we first started, we were doing just residential. In fact, the first site is my house or was my old house, right? It was interesting. It was a 7 kW system. The second site was 125 kW. The third site was 1 MW.
This demonstrated to us early on that we had the right pieces of the puzzle and the right technology to kinda grow. We went through orders of magnitude within a very, very short period of time in terms of the size of the sites that we were taking care of. Very excited about that. Just very, very excited. At the same time, you know, just a year after we started getting revenue in 2009, in 2010, we were lucky enough to attract an anchor customer called Morgan Stanley. Morgan Stanley saw some things in us that we just didn't understand yet, and they standardized on us of all crazy things.
You know, that was interesting from our perspective because we were just trying to wrap our brains around what was the value of the standardization. Honestly, what it did is it reduced their cost of human capital to manage all these systems. Instead of having 10 different systems to go into, they had 1. They had 1 runbook, right? That allowed them to really maximize the value of their systems by minimizing the human cost of managing assets. In 2017, we attracted a private equity firm. We looked at that 'cause we looked inside ourselves and said, you know, "How do we continue to grow the company?" We said, "We have skills, and we got gaps." In 2017, private equity firm came. You know, we attracted a private equity firm.
In 2018, we acquired 3 companies in the course of 90 days. It was a little harrowing, a little difficult. W hat those 3 companies did for us was kind of interesting. What they did is they expanded our market coverage as well as our geographic coverage. You know, 1 company we acquired was heading into bankruptcy. It was a customer acquisition, right? Pretty straightforward. A second company we acquired had an office in Berlin and had a utility expertise. With that office in Berlin, we also acquired an office in Tokyo. We had now, with the second acquisition, you know, a more of a global coverage into key markets which we thought would lead the world in renewable adoption.
Third acquisition that we did further you know kind of strengthened our case and strengthened our sales here domestically but also gave us an office in Delhi, India. Now the AlsoEnergy team could which was located here in States which was located in Berlin and which was located in Delhi and Tokyo could honestly say the sun never sets you know on AlsoEnergy. That was kind of an interesting comment and you know it just made us very proud of what we were able to do. You know what happens here you can see rapidly then we grew in terms of asset under management in terms of skills and capabilities in terms of segment coverage and in addition in terms of geographic coverage. What did that lead to right?
Strategy is great. I love execution, right? I do. For us, what happened now is, you know, we have customers now that have standardized on our platform, just like Morgan Stanley, but now we have a better understanding of why. If you look at the data here or the information here, it shows, hey, 13 of the top 15 U.S. C&I commercial and industrial asset owners have standardized on PowerTrack, our platform. 9 of the top 10 community solar asset owners have standardized on PowerTrack. To me, you know, our customers, right? This adoption is evidence of the market-leading capabilities. You know, we have companies like NextEra, Clearway and Nexamp that have standardized on our product. The great thing about that is, from my selfish perspective, you know, as our customers grow, we grow with them. Right.
We grow site at a time. Some of these customers, they have not 1 site or 5 sites or 10, but they have 100 or 1,000 sites. That's where the real value of the platform that we provide comes in. The interesting thing about this also, when we look at standardization, is it really creates what I'm gonna call a flow business, where we do many, many transactions. We'll do several thousand transactions and be responsible for several thousand sites on an annual basis. It's a nice flow business. It runs. We're always actually, you know, looking for new customers, but we have a great base, a great and growing base with a continued increase in the number of sites we have under management.
Again, if we look at it, you say, "Hey, you know, so Bob, you created the system," you said, "Hey, I've got this." People have standardized. You may be wondering now why, right? What are the key customer value propositions? I kind of touched on them a little bit, but I'm gonna repeat them here. I always think about renewable sites, these solar PV sites with storage, sites with solar and/or however it works, right? These are simply investments, right? They're simply investments with elements of risk, liquidity and yield. You know, quite honestly, when you have 10 or 100 or 1,000 of these sites, you can't manage them with spreadsheets, SharePoint, with any of these things. You have to go to a platform, right?
The value we bring to our customers is that we have a platform, we have expertise. We're subject matter experts in what we're doing. They can buy that subject matter expertise as part of a SaaS component. What we do then with our SaaS product, you know, is we maximize the yield, we maximize the revenue from the site. That's done because in near real-time, we're able to report system outages or system underperformance. You're able to understand, you know, how your systems are performing against what they should be doing, right? Predicted versus actual. Predicted versus production, right? These are important ideas. At the same time, we decrease the cost of operations and maintenance, right? This is the human element.
The reason and the way we do that is we're able to pinpoint the problems on the site, so when truck rolls occur or when corrective actions need to be made, people know what to do, right? You don't want to do this twice or 3x . You want to do it once, right? For us, it's all about increasing the revenue and decreasing the cost, right? That's the things that we do. Finally, as we talk about decreasing the cost, now, since you have a single pane of glass, a single asset performance management portal, right, you as an asset manager can run standardized reports and KPIs across your fleet, understand how they are performing against their prediction, how they're performing against each other, how they're performing against things that could be close by.
Quite honestly, you get this great 360-degree view of your asset, right? Again, it's all about how do you increase that yield? How do you increase the value of those assets? Again, that's precisely what our platform is designed to do. Where does that take us? I said early on, right, in this presentation that the vision that I told our folks was that in the future, all generation on the grid was gonna be renewable, driven by cost and resilience. Honestly, AlsoEnergy could not deliver the promise of resilience in the absence of storage, right? We couldn't do it. We could never achieve the full vision statement. That's what we're able to do today, right?
With Stem Athena, PowerTrack, the piece of the puzzle you have here in front of you today, you know, we're able to further maximize. This is what Rakesh said. We can further maximize the return of these renewable investments. We can further reduce the risk associated with the investments and really drive the adoption. It's an incredible combination, right? It's a vision I'm happy and proud to tell. I appreciate, again, you listening to the presentation and hope to be able to answer any questions later on. Thank you. Now I'd like to introduce Larsh Johnson. Larsh is our Chief Technology Officer.
Please let me know if you're not able to hear me. I will be going through the program here and hopefully give you a sense of the differentiation of Athena, introduce you to our technical talent. We have a number of them who are gonna be joining us a little later. Give you a demonstration and show you what's happening with our platform. You know, as Bob mentioned, we have a variety of different things that the platform does. First and foremost, it'll optimize the economic value of assets through both the effective and intelligent dispatch of flexible resources like storage, and it will improve the overall asset performance through the lifetime management of the solution.
We'll be covering that in a number of these slides. I'll start off with that because that is a key differentiating factor, is to have that combined unified platform that services those 2 key values. If you followed Stem, you know that this energy transition is the problem that we're solving. We believe that smart energy storage is the key to this solution. What's been happening with the you know, the various changes in the markets and regulations is that the value propositions for storage are getting richer, they're getting more complex. The same thing is happening in the solar industry as we see the attach rate of solar and storage you know in new project development continuing to increase.
With the success of these programs, these value models have gotten more complex and we've been able to, you know, essentially envision this platform that allows us to make this work. The problem with complexity, though, is that it's sand in the gears, right? The sand in the gears is the real problem that we'll end up solving with software through automation and the effective use of technology. To accelerate this transition, we need to make smart energy simpler. You're gonna hear the simplifying theme throughout the presentation. Athena is an enterprise platform developed specifically to make solar and clean energy storage simpler and more scalable to affect the energy transition.
Let me introduce some of the people that are developing this and talk about our team. This Athena and PowerTrack is a unified platform. It's built over 10 years with multiple different generations. The operating experience in the school of hard knocks has been a big part of making sure that we deliver the most robust platform in the industry. Today, over a third of the company is devoted to this endeavor. 200 professionals with deep experience in all aspects of software as a service, data science controls, and industrial IoT.
If you've heard from John, the experience of this team and the methodologies of our development, the technical capabilities have been ranked best in class across the digital leaders, let alone our industry peers. It's a super accomplishment, very proud of this team and the work that they've been doing. Getting there has taken some time. When I started with Stem almost 7 years ago, we were delivering software releases on a quarterly basis. We even named them after BART stations in San Francisco, as a train was leaving the station going into the next stop. It took us 90 days to get that process completed and get that software out the door.
Best-in-breed software today, we're delivering software on a daily basis in incremental releases that allow us to continually innovate and update software, all while we're continuing to operate a 24/7 platform that's today managing over 500,000 end devices, 32 GW of solar and 2 GW of storage assets together, having accumulated over 1 billion hours of operating time. It's truly remarkable that we're able to continue to operate in this very fast-paced implementation process. These best-in-class ratings also apply to our product strategy, as you saw on the chart.
Our agile development methods where our experienced commercial and product teams evaluate the needs and opportunities with the grid, incorporate our voice of customer and prioritize and develop and deliver all of these Athena innovative solutions. Very proud of this team and want to continue to highlight the successes that they've developed here. Let's talk a little bit about our platform and what makes up the platform that we offer here. The software is key. It's first and foremost probably our largest investment in terms of our people. This open enterprise platform is today supporting a growing number of software applications that simplify the full life cycle of our clean energy assets. To do that, we have a second part of the solution, which is the edge solutions.
While we often talk about hardware and you think about batteries or solar panels, there's an important part of the overall operating plant that are the controls, communications, and the compute power necessary at that local site in order to provide the instrumentation to measure, manage, and control those assets. That control equipment, that software and equipment deployed at each site, we call that the Edge platform. AlsoEnergy has been a leading provider of Edge platforms for the solar monitoring business and controls. What we've been able to do is to unify the platforms with both solar and energy storage and provide a single unified approach to instrumenting and controlling these large assets and the equipment that we operate.
When we talk about our Edge platform, it is a combination of our proprietary software extending Athena at the edge and the equipment that is necessary to run the compute and communications to monitor and control each of those assets at the site. Third, we have an extremely strong group of expert services personnel that provide subject matter expertise across a wide range of applications and capabilities that are very important to the overall lifecycle management and simplifying the process with our customers, starting at the conception of a project all the way through to the operating metrics that achieve the value of the asset over time.
Those 3 elements of our platform are all key to our offering and a big part of what you'll continue to see us delivering to the customers. Together, they form the trusted enterprise platform that unlocks the flexibility across the clean energy value chain. This is an important set of words. It defines what we deliver. It defines our commitment to our customers. That is starting with the idea that it's a trusted platform. That's rooted in the performance and the metrics that you've seen, the deliverables that we have for our customers, and the value that they get by working with Stem. The flexibility is in the 13 different value streams that you've heard us talk about.
Today, we have a new icon, so we're gonna throw a little curveball at you, but the summary of those 13 value streams is shown around the rim of this particular wheel. This is essentially the flexibility to, you know, provide full solar optimization using storage to achieve utility bill savings for C&I customers behind the meter, to participate in wholesale markets and earn grid revenues from dispatchable utility services. These are provided both in front of and behind the meter. As you heard also from Prakesh, to address some of the GHG objectives that customers have, whether it's through incentive programs or, you know, internal ESG objectives that many of our customers have in terms of their sustainability management. Finally, we provide resiliency and backup power.
As the people in Florida are probably finding out right now, you know, that can be very important in some of these different weather conditions. Flexibility is not just for the initial economics, but for future-proofing these assets. As we see the landscape changing, markets and regulations evolving, you know, it's important for our customers to know that even though they may not be able to take advantage of one of these values right now, in the future, they may be able to, or it may become more important to their initial than it may be initially when we start the project with them. Then finally, enterprise. You've heard Bob talk about how customers are standardizing on a platform like PowerTrack.
We see the same application, you know, being able to provide, you know, a unified performance management window across behind the meter, front of the meter, storage plus solar alone, and then any type of different OEM equipment that needs to be integrated on a single platform is an incredible value to our customers who are continuing to grow and invest in these assets over time. That is a unique breadth and depth of clean energy solutions, value streams, and asset performance management that we feel is highly differentiated in our offering today. We're gonna unpack that software a bit, just a little bit and talk a little bit more about the 2 categories we've been mentioning here, economic optimization and the asset performance. First, Athena automates dispatching with flexible energy systems, right, to achieve that maximum economic value.
How does that happen? Now, the first step is we need to predict the future. We need to both predict what's going to happen. We need to understand how much solar generation we're gonna have available. These are at every particular site that we're operating. We need to predict that solar generation. We need to predict what the market prices will be. We need to predict a customer load if we're gonna be offsetting a customer load and optimizing for a behind-the-meter facility.
We need to run that optimization, which oftentimes takes millions of scenarios in order to achieve the best fit and look at the landscape of that particular site over a multi-day period, determining if the energy available is best to use today or best to save and use tomorrow when market prices may be higher or opportunities may be better. We do that over a multiple day, you know, scenario. We're doing this continuously, literally running millions of scenarios every hour. Finally, we have to dispatch the asset.
Once we determine what the economic, you know, ideal operating plan is, we need to actually create that realize that economic plan either through bidding into a wholesale market and clearing at prices that will affect the operation of the system the way we want or by directly dispatching that asset for a particular customer opportunity. This automated and very scalable prediction optimize and dispatch process run continuously, nonstop, 24/7, supporting the economic dispatch of these flexible assets. This is a fully automated process, lights out. There's no humans involved in this loop. Once we set it up, it just keeps running. Second, towards the point of asset performance management, the whole time we're doing this, we're continually getting streaming data from these sites. Data that can range from every second to every few minutes.
It's coming in order to make sure that we have complete situational awareness for all assets, and that Athena can determine if there's anything that needs to be acted on. Corrective actions such as, you know, remote automated solutions as well as corrective actions that may involve field work dispatch. Athena automates this process, processing terabytes of data every day and determining where there are issues, and then triggering the appropriate workflows to either automatically resolve an issue, informing the proper people, and then in some cases, being able to initiate a workflow that results in a truck roll in order to do some work at the site.
Delivering both of these again on a unified platform, we think is a entirely differentiated offer that increases yield, reduces risk, and improves the operating efficiency of these clean energy assets. This is just incredibly important to achieve our mission of you know accelerating the clean energy transition. Because the cost of these operations, the economic value, is what underwrites the ability to continue to grow this clean energy portfolio. We're gonna dig in 1 more level and talk a little bit about the automated AI operations. It operates on several levels. You'll hear us talk about you know 1-second data. We're also talking about 1 -second and even sub-second control that has to happen.
A lot of this is affected directly at the edge because of the latency that has to be very low for some of these grid-related services. That's these kinds of cadences are effectively, you know, what lead us into the performance that we've been citing. 96% perfect foresight in terms of the ISO New England market revenues and overperforming in the greenhouse gas goals at C&I sites in California. Again, every second, these solutions are operating at the edge. They're acquiring site-specific data, conveying that real-time asset condition, generation, local weather status, and they're making control decisions that can respond in ms. At least every hour, Athena is executing the digital model of each and every site.
This is being done in the cloud. It's leveraging that real-time data that's streaming to the cloud from every one of these sites. These individual models of each site effectively will generate predictions for solar generation, site loads, energy market prices. Then those predictions are used in these millions of operating scenarios that are evaluated in order to determine the best way forward for that individual site. Again, that's being optimized across a multi-day period, typically, you know, 3 to 4 days.
On a daily basis, we're evaluating the key performance indicators, the KPIs that really determine how we're performing either against a greenhouse gas goal, against an incentive goal, against battery longevity and performance objectives for the battery and warranty and so forth, and then determining, you know, what that ongoing operating plan should be in relation to those maybe longer horizon metrics that are also part of the optimization process. Very interestingly, quite important is on a monthly basis, you know, our automated AI operations will evaluate the models that are used for each of these forecasts for each and every site.
Every site may have multiple forecasts, you know, streams that are being run, and every one of those may have a model that's selected specifically based on, you know, a model evaluation process that runs automatically every month in order to adapt to longer-changing conditions. That automated model selection can look at a deep neural net, a gradient boosted network, a gradient boosted forecast, you know, a random forest approach. All of those different approaches are evaluated in order to say what's the best one for this particular site. Then that's selected for the next operating period.
Athena is automatically doing this in the background, evaluating the best model approach, the set of hyperparameters that control that model, and then being able to say, "This is how we should proceed." That AI automation for that process is something we think is also a unique and differentiated value in terms of scaling and operating these systems, adapting to conditions without human intervention. Of course, you know, our data scientists, as you'll hear later, they love to get hands-on with data, and so they get very deep into some of the operational reviews and so forth in order to say, "Well, how are things going?
What are some of the benefits of this particular model approach or that model approach, and how do we improve that?" Typically, we feed that into our product development process, but then we can very rapidly deploy software, as you heard, in order to adapt to those changes as well. This is part of the overall platform. Once we've looked at that, you know, those core capabilities, we think about, well, now, how are we gonna use these? What are some of the applications that can take advantage of these capabilities? We'll talk about 3 of them here. This is 3 out of a longer list, and you're gonna see some of these demonstrated today. The starting point for this is to understand what's the
What are the customer personas that need some of the value that we can create through these Athena processes? That starts in the beginning of a project when we need to evaluate the economic returns and do the simulations in order to say, "What's the best battery? What's the best size? How are we going to model that with different value streams over time?" The Analyzer product is used by the developers in those particular cases. As you heard from Bob, O&M providers are using PowerTrack in order to monitor and determine when the assets need some sort of repair and determine what the asset performance is against expected capabilities, and then indeed being able to affect that corrective action when needed.
Asset owners, who are a key target for us in terms of our enterprise software platform, are the ones who own these assets over time, sometimes 10 to 20 years, operating objectives that they have and being able to use our software for ongoing economic, you know, performance evaluation, understanding where it is relationship to the expected, and then deciding, you know, if they wanna change any kind of operating strategies that may affect the future economic value of that particular site. As we do all this, we're using a common enterprise platform that leverages a common data plane that ensures consistency in the reviews of the asset and the performance against any one of these vantage points from any one of these vantage points.
I've described some of the core Athena services that can be leveraged in multiple ways and a few of our applications that demonstrate, you know, how we're going to be serving different parties throughout the ecosystem. If we look at this, you know, and talk a little bit just about the core platform itself, we think about really 3 different areas of capabilities within this software infrastructure. We call these core, common, and distinct. If we think about the value of a platform, the value of a platform is when you can have core and common services that allow you to rapidly build new individual applications without having to replicate that underpinning.
The infrastructure for data acquisition, the infrastructure for data quality and data management, the infrastructure for doing forecasting and optimization, all of these services become part of that building block that allow us to assemble applications and achieve a great deal of software reuse and what you would call operating leverage as we attempt to expand into new markets and do new things, you know, with new use cases and even new sectors.
Examples of this include the ability to move into EV charging space because the data acquisition infrastructure and the data modeling for achieving telemetry from EV solutions is in place, and it's extensible based on what we have in our core data acquisition services that today process over a terabyte of data every day from a whole variety of different endpoint devices. As we think about the core services, I had mentioned forecasting and optimization. As you think about what we're doing throughout the life cycle of the project, you know, before we start a project, we'll analyze the project performance using the same optimization process we use in production operation, where we're running controls, we're using the same optimization service.
When we look at retrospective review of performance, we'll compare that operating performance against a perfect foresight using the same optimization process. Those components are being reused in multiple different ways to provide applications that cover the full life cycle of a project for different parts of the customer ecosystem. We're gonna move on to the third part of our platform, and this is our services. As much as we automate things, there's still some things that humans need to do. The software platform it's got a terrific range of capabilities, and our customers value that, but they also value the subject matter expertise that we see from a team of experts that, you know, have been working with solar and storage for many years. Storage is much more, let's say, nuanced today.
It's not as far down the learning curve as solar. There's a lot of product iteration, there's a lot of new value streams, there's a lot of variability in the way storage systems get used. A lot of our customers are really trying to come up that learning curve themselves, and in doing so, they're willing to pay us, you know, for these kinds of services. We've proven our expert services and bundled offerings, and now we're starting to announce services that will be available on an unbundled basis as, you know, in a number of these different areas.
With that, we think we can capture the full value of the subject matter experts that we have on staff to help in the design process, to be able to bring Athena-ready sites into a commercial operation, simplifying the process for both the developer and the engineering and construction teams. We think we'll be able to provide network services for those folks who don't have remote operation centers. Our remote operations team today provides full life cycle management, monitoring, and managing the overall sites, coordinating dispatching to multiple different OEMs and field service providers.
Very importantly, in our program operations team, we provide a sort of almost an owner's representative or an agent that enables that team to enroll systems and assets into wholesale markets, into utility programs, and provide all of that back office support in order to get to the value streams that are part of wholesale market and grid service participation. Taken together, we really have a full set of, you know, software, edge solutions, and program services and operating services with subject matter experts that can really simplify the process of getting a scalable approach to deploying clean energy assets. The question is, well, what does winning look like? With this extensive suite of capabilities, there was a lot of different characteristics that I think are important here.
The interesting thing that's going on with the passage of the IRA is how to think about solar plus storage combination. We're gonna focus a little bit on, you know, winning with the hybrid systems where you have both solar and storage. In particular, potentially retrofitting storage to that 40,000 installed base that you heard about with AlsoEnergy PowerTrack system. We'll start with one of the sites that has solar and storage in Massachusetts. We've been operating this site delivering a full suite of capabilities that automatically, continuously co-optimize 7 different value streams. We're bidding hybrid solar and storage into the ISO New England market. PowerTrack is at the site providing a complete view of the asset, the solar asset and on-site weather conditions.
This combination of economic optimization asset management, you know, has resulted in revenues exceeding the project pro forma by over 46%. This demonstrates the differentiated capability to operate and monetize these 7 different value streams, you know, in grid services and wholesale market opportunities in a front-of-the-meter site operating with the combined services of the Athena platform and the AlsoEnergy PowerTrack application. Moving behind the meter with a continuing growth in solar and the inevitable duck curve that's occurring in multiple different markets already, we're gonna see utility tariffs and wholesale market prices and incentives that drive a lot of the economics, as you see in this picture, you know, in other jurisdictions as well.
The economics for storage are gonna shift demand to optimize solar, and then to be providing fast-acting services for, you know, grid operators in order to stabilize the grid with increasing penetration of solar. In this example with solar energy, the customer had solar energy. We significantly increased their savings, adding storage as Athena monetizes these value streams, and this is optimizing against that customer's forecasted load. It's optimizing against utility tariffs and market prices and then also again, forecasting solar generation in order to do so. That optimization forecasting loop that we talked about is in full swing here, optimizing this value for this site, delivering these kinds of returns for the customer.
I think as an example picture of what we can see as we look at that installed base of solar that now with the IRA passage is likely to benefit from attached storage. Just as a reminder, you know, this is the bread and butter that Athena grew up on in the California market where these conditions were all present. We've been able to continue to establish our leadership. You've heard us talk about the project that we took over in 2020, where we've been able to demonstrate superior returns for the asset owners and the customers on the order of 30%. That's another, you know, again, proof point of our differentiated performance in this particular market segment.
We'll see that Athena's advantages will continue to play out in California. Not only the IRA, but the rollout of the FERC Order No. 2222 ruling that we now see, you know, taking shape in multiple different markets where wholesale market participation is starting to become more tractable and economically valuable to behind the meter assets. One of the other dynamics that's going to come in here is the merger and the co-optimization across what we're already doing today in multiple markets with front of the meter projects moving into the behind the meter applications as well. Again, bringing back the differentiation of Athena with this broader suite of capabilities across a broad market segments, and being able to tie that back under a single platform into a unique value proposition.
Another aspect of winning is the idea that we expand our partner ecosystem. We've announced a relationship with EV charging leader InCharge Energy, and it's just another example of a expanding ecosystem where we're adding value, taking adjacent capabilities and being able to integrate that with the capabilities that we offer. In this case, you know, being able to, as they say, fuel the fleets from the sun, optimize minimize the grid impact and then providing resiliency so the customers can continue to charge their fleet vehicles even if the grid may be unreliable at that time.
As we look at these applications and, you know, some of the partners in our ecosystem look at these applications, they see what Athena and Stem is doing with utility tariffs, market pricing, the complexities of that, the complexity of integration into wholesale markets and grid services programs, and they see that we're already doing that. Then they see that we're also, you know, able to optimize for greenhouse gas and then offer resiliency. They say, "Well, this is a no-brainer. We have to work with Stem, you know, to achieve these goals in addition to our fleet management and fleet operations goals for EV charging." Taken together, we offer a solution that I think will mirror some other electrification initiatives that we'll see coming as well.
It is definitely being led, as you heard from Prakesh, by the fleet electrification and EV charging. Finally, you heard Bob mention the standardization on the PowerTrack platform, and we see winning as owners standardizing on the Athena platform. All of that means the entire combination of the services that we offer for economic optimization, the services we offer for asset performance management with the PowerTrack application, and the services we offer fundamentally as a company. We've seen that PowerTrack has been able to achieve that status, and we think that one of the ways that we'll continue to drive that for the broader platform is to focus on applications that address the full ecosystem.
While owners may be the long-term target for our sales efforts, we're also able to sell our capabilities, some of them I've described, into all different parts of the ecosystem. As I mentioned, developers wanna understand the project economics up front. They will be looking at the Athena Analyzer and the capability to model and assess the value of projects and different value streams, look at pro formas that may do the what if on what happens if policies change, and be able to create value there and simplify that process of achieving a bankable solution that will have great value to the asset owners that ultimately acquire those solutions. The EPCs, the engineering and construction companies, will see simpler deployments.
The unified edge that I mentioned will allow them to more effectively model the control and bid the cost of implementing controls for a project site that they're responsible for, and to have services that can help them, you know, commission sites and make them operational before that commercial operation date, where they may be transferred over to the operating partners and the owners. O&M service providers will be key long-term users of the Athena platform.
They'll be using it to monitor the systems that they're responsible for and maintaining, and they will be able to use that to initiate the workflows that will ultimately help drive the effective use of resources to minimize the operating costs and to make sure they deliver the best value and best performance to the owners of those systems. In full circle, of course, owners will have a trusted enterprise platform.
It'll offer a single pane of glass that covers both behind the meter and front of the meter facilities with a mix of clean energy assets across multiple markets, use cases, and OEMs, and provides integrated economic optimization and asset performance management. That's a mouthful, but that's what the platform promise is to be able to provide that kind of capability with 1 stop, with 1 company who can support that over time. When compared to multiple point solutions, our unified platform and Athena offerings are differentiated in the industry, and the ability to sell solutions into this entire ecosystem will make smart energy simpler for all. Now I think we're gonna take a break.
Okay. L et's take a break for about 10 minutes or so. Come back around 10:30, and we'll restart with a software demo. 10 minutes and be back in the room. Restrooms are around the outside if you go outside there, and we'll be back. Thanks. All right, let's. If everybody can sit down, we're running a little over. We wanna stay on track. If everybody can take your seats, please, we're gonna get restarted here with the software demo. What's that? I'll have a bell. I'll just go grab people.
Hey, everybody. Welcome back. We're gonna go ahead and get started on the next part of the presentation here. We need the little ding, ding. The chime calling next session. I wanted to introduce you know some of the team here, and in particular Bryan Ho, who's gonna lead off the demonstration of our software products and applications and talk a little bit about the value proposition to the key customer users and so forth. I wanna brag a little bit about Bryan and his background. First of all, he's over-educated, like many of our folks, with a Harvard undergraduate and a PhD from MIT. I'm very jealous on that front.
Then he founded an energy storage company and then was the head of product at energy storage for NRG for, what was it? 6 years?
6 years.
6 years. Obviously very, very well trained, very well versed in the economics and business value of storage, and he's here to demonstrate some of the capabilities we have with the Athena platform and applications. With that, you're on.
All right. Well, thank you for that very kind intro, Larsh.
Oh, well deserved.
Over the next 15 minutes or so, I'll take you through a tour of our Athena apps. As Larsh mentioned, our apps sit at the top of the Athena platform and provide that window to connect our asset owners, our customers, with their assets. As I walk through the apps, I'll do my best to highlight some of the powerful data and analytics going on under the hood that power those apps. We see here we've got Explorer, which is our asset owner, asset manager facing application that provides high-level insights on financial and economic performance of the assets. We've got PowerTrack, which provides asset performance management capabilities across asset types, really targeted towards asset managers, operations managers, and field technicians. We've got Analyzer, which is developer-facing, which provides analytics on pro forma revenue expectations on projects.
Finally, we've got Supervisor, which is a network operator-facing application that provides a bit more detailed views of the telemetry and operations of the asset. The Athena sort of 1 platform idea is something I'll emphasize as a theme and come back to as I go through these applications. It's the idea that across our applications and across our capabilities, we offer asset owners a one-stop-shop regardless of asset class, whether that be solar, storage, or now electric charging. Across asset scale, whether that be C&I or utility scale, and across asset activity, whether that be asset performance management or economic optimization. We'll start with PowerTrack. Diving right in here. When you log into PowerTrack, you're greeted at a portfolio-level summary of key performance indicators around your assets that constitute that portfolio.
Here we see a list of sites within the portfolio and a completely configurable table of KPIs. We can bring in KPIs like availability, which could help an asset owner quickly identify compliance with contracted targets. We can bring in a summary of any active alerts, and by hovering over, we can see details on those alerts. We can bring in reminders, production numbers against rolling averages. This whole view is deeply customizable as we understand that asset owners and users will have their particular list and priorities amongst indicators that they'll use to gauge portfolio health. Another important portfolio-level view is gonna be the event manager. Here we bring up across the portfolio active events, typically faults and alerts that are actionable for the asset manager and their agents.
Actionable is an important word here because here I'm going to unhide some suppressed alerts. With this kind of view, nuisance alerting is a big problem. If you flood a user with information that is not actionable, they lose the forest from the trees, they lose the valuable insights among the noise, and it reduces the value of the tool entirely. What's going on under the hood is, you know, with all the information, all the false alerts coming in, analytics are being applied to those to categorize them, nest them, and determine which ones are actionable.
That at the end of the day, we have a very clean list of items that an asset manager or operator really needs to concern themselves with and can take the remedial steps to bring their portfolio back up to a state of operating efficiency. We click into any of these particular items, we can immediately acknowledge them. We can also start creating work orders. The idea of Athena as an open platform means that we interact with third-party systems to enable seamless workflows that provide Athena as a headend to connect in with our customized enterprise systems to again enable to get a job done. Now, in addition to these portfolio-level views, once you start talking about operating managers, operations managers and field technicians, you might wanna go deeper into a specific site.
When we look at site-level information, we get immediate real-time data across solar and storage assets. We get data and configurations on all the hardware on site, as well as production, time series data. If a site is of particular interest and there is an issue that needs to be, you know, taken a deeper look at, we can build custom charts. You see here that we are bringing in data from across a large number of devices on site, and that's really another sort of key value of PowerTrack, is its open configurability and compatibility with the hardware out there. That enables asset owners to bring the hardware they choose to the projects and connect up into PowerTrack with what Bob mentioned, this concept of standardization across a portfolio.
Once it's in PowerTrack, everything looks like all, you know, all weather stations look like weather stations. All inverters look like inverters. We standardize on that data plane. If we wanted to look at inverter data, we can look at specific inverter data down to highly granular information. For example, data coming off of the DC bus of an inverter. This would provide the tools for a technician to really get into root cause diagnostics, seeing which inverter is deviating from the norm or what that behavior looks like, and, you know, roll the appropriate truck. Now, once we get past sort of the, you know, asset performance management aspect. You know, the other side of the house, you know, might be interested in for these sites, how is it performing from a financial perspective? What kinds of revenues is it generating?
For that, we'll hop over across to the Explorer app. The Explorer app will look at a similar solar plus storage site. As we land on that site, we're immediately greeted with a summary of revenue performance on that site. You can look at both a storage view as well as a combined solar and storage view. On this page, you get a monthly view by utility billing cycle of how much the facility has saved on their bill with the deployment of solar and storage. We can see some very high-level breakdowns of where those savings came from on a billing period view. If we dive into the data any given month, we can get an even more granular breakdown.
Here on this site, we're looking at a month's view of data, both solar production, original building load, and net building load after the contribution of storage and solar. What you'll see here is the storage is working hard to keep those peaks of the facility at bay, thereby mitigating their demand charges, which are a common form of utility charge amongst C&I customers. You see the solar carving out large amounts of load during the day, reducing their energy charges. Combined through the breakout of this monthly bill into its subcomponents, we can find that of the original $28,000 monthly bill, solar and storage was able to reduce the bill down to a little over $15,000.
For those asset managers who are particularly interested in what drives the behavior of the savings, we can go to views that provide some of the input signals into our optimization plotted against the time series data. Here, this site is optimizing for utility bill savings as well as greenhouse gas reductions. We see that same operating data. Against it, we can now see how the retail tariff charges are changing with time, typically having a peak period, a partial peak shoulder peak period, and an off-peak period, and how the marginal emissions on the grid are fluctuating over time. It's co-optimizing against saving on the utility bill and again, generating reductions in greenhouse gas emissions for the facility operations to finally arrive at the ultimate operations of the site. It's worth kind of noting here what's going on, you know, under the hood, right?
To accomplish this, we have an up-to-date high-fidelity database of tariffs. You know, you need to have good tariff information to operate properly. We've got advanced forecasting models that allow us to forecast building load to best understand how we can use the limited energy in a battery to maximal effect. We're operating through a model predictive controls loop, as Larsh mentioned, updating our forecast and optimizations every hour to execute on the optimal dispatch strategy of that storage system. Now, once we actually dispatch a storage system, we then take in the meter data. Again, we reconcile it against utility bills as well as the retail tariff information to then come up with those summary savings metrics.
In these situations, Athena sits really at kind of the center of intelligence for an entire facility, which just allows us then to summarize for a facility owner key information about their facility, how their peak demand is evolving over time with and without the aid of solar and storage, as well as how their energy consumption is evolving over time with and without storage. Into this sort of center of intelligence for the site, we're now adding new capabilities, and as Larsh mentioned, EV charging is our next step, our next asset class. As we diversify now to solar, storage, and EV charging, we're updating our Explorer to now bring more relevant information to the user immediately in the form of customizable KPI tiles.
Here we immediately bring to the attention of the asset owner, the savings generated, the solar generation on site, and quantify the greenhouse gas reductions. We also provide a unique tile for them to get clear information about how their EV charger status is in terms of what power is being discharged by the EV chargers, what the availability is, and then again, with this concept of an open platform, allowing them to link in their charge management system natively through Athena. In addition, we're creating new views, particularly around greenhouse gas emissions that'll really allow facilities to track their emissions and the emissions reductions they're achieving over time. We sent a couple of slides looking at what, you know, solar and storage is doing for the C&I customer.
We're now gonna switch tracks a bit and go on to our Supervisor app to look at the wholesale market participation use case. We're taking a look at a solar plus storage site in Massachusetts, participating in the ISO New England market for energy and ancillary services. We bring in real-time data into Supervisor, and again, this is the network operator-facing applications. As we chart historical data, I'll give you a bit of a background on what's going on on-site. This site, again, in Massachusetts, is part of the Massachusetts SMART program, which has incentivized many new solar facilities to co-locate with storage. The storage on-site is participating in the wholesale power markets independently, but there are operating constraints placed on that solar asset.
It will claim the storage asset claims solar ITC, and therefore must charge off of solar energy. The project also has a limited interconnection capacity, so the storage cannot discharge concurrently with solar. Under these operating constraints, Athena is optimizing the operations and dispatch of the storage and its bidding behavior to earn the maximum amount of revenue in the ISO New England wholesale power market through both the day-ahead and real-time energy markets, as well as the forward reserves, real-time reserves, and real-time frequency regulation markets. The activity in those and the dispatch by ISO New England in those markets are what you see here as the purple.
Again, looking under the hood a bit, in order to execute on this use case and sort of derive revenue from that ISO New England markets, Athena is forecasting prices for each of those market products. It's forecasting the intermittent solar generation on-site to understand what it can bid in the future to not violate these constraints of solar ITC charging or interconnection. It is then taking on other site-level constraints, related to hardware configuration, other programmatic rules, from participating in the SMART program, and then optimizing all that to generate bids. Prices and quantities are submitted to ISO New England. When those bids are cleared, the dispatch signals are routed back out to the asset, battery dispatches.
On the back end, we provide services to our customers and asset owners to summarize settlement data and report on performance. We'll take a look a little bit here at the bidding behavior. We can go in and look at, for example, what the battery is doing in the real-time energy markets. We've got on the top graph, historical prices, and on the bottom graph, we see a combination of data related to our forecasted prices, as well as the quantities that we're bidding in to buy and sell energy, and the prices associated with those quantities that we're bidding and selling energy at. The peakiness you see in the hashed lines are the hours at which we're choosing to participate and buy or sell energy in that real-time market.
The solid curves are the prices associated with those quantities at which we are buying and selling energy. The dashed curve is our forecast for the energy price. We'll generally, obviously, buy energy when prices are low, sell energy when prices are high. We'll also transact in order to maintain the state of charge to posture for our participation in ancillary services, particularly reserves and frequency regulation. We can take a look now over at what we're doing in frequency regulation, specifically. Here we see within the green and red hashed envelope is the capacity we're offering into the market for frequency regulation. We'll offer in when frequency regulation prices are high or during even moderate priced hours, if the battery is relatively idle, we will sort of monetize those idle hours with participation in the regulation market.
The importance and kind of the value of Athena as an intelligent platform is again co-optimizing against both day-ahead real-time energy and ancillary service markets. By bringing that full bundling together and optimizing across those, we provide the optimal revenue outcome for asset owners and sort of the greatest uplift we can possibly provide over manual control of these assets. With that, I'm going to go ahead and return now back to our Athena landing page. Again, you know, Athena is that 1 platform where our users, regardless of asset class, asset scale, and asset activity, can come and connect with their assets. All right. Well, thank you very much. We're going to put on a short video now as we transition into our panel discussion.
Okay. Just going to do a quick testimonial video.
My name is David Arfin. I'm co-founder and CEO of NineDot Energy. NineDot Energy develops and operates battery storage sites on a community level in New York. It's a great time to be in clean energy. The pricing signals, the tariffs, the incentives, the public will to do something about climate change are all tailwinds for us to be able to build a sustainable, important business that's going to service the New York City grid for decades to come. Stem makes it simple for us. They help us model sites, procure the battery hardware, and then help us operate the sites with the Athena software platform so that we can optimize these sites over time. We have a very strong relationship. It goes to calls, to texting, to meals together.
We solve problems together, and we do that not just with the account representatives, but we also do this with the team in contracting and with the teams in design and the teams in software. We're going to be able to do some really cool things coming up in terms of how clean the power that charges our batteries are. Stem and their software platform, the Athena platform, enables us to buy electricity at just the right time from the right source. As we expand our portfolio, we expect the flexibility of Stem solutions will ensure that our projects are successful now and throughout their lifetime.
My name is Michael Smith. I'm the CEO of ForeFront Power. ForeFront Power develops solar and energy storage systems for commercial and industrial customers, as well as for community solar applications. The solar EV and storage landscape is evolving very quickly. A lot of customers are choosing to install on-site generation to electrify their fleets, and we feel like we're right in the middle of that. We've been a long time partner with Stem for our energy storage needs, as well as with AlsoEnergy. We feel like they provide the right suite of products and services for us to help us serve our customers. A more unified platform for energy optimization is critical for large customers.
As more customers install more on-site generation and storage, it's really important that they're able to optimize the entire suite of products and services they have. It's important both for them as the customer and us as the provider. As we continue to scale our solar and energy storage business nationwide, Stem has become one of our preferred partners to fuel growth. We're looking forward to the integrated view of our assets and superior optimization that Athena and PowerTrack will deliver to our customers.
Thank you.
I'm back to introduce our leadership panel. We're gonna have some great discussion with the team here. I'd like to introduce the man who's gonna MC this process, Albert Hofeldt. He's our new SVP of Technology. Albert joined us in September and comes to us with a rich background in energy markets, IoT devices. One thing that was particularly interesting from my perspective was his experience with financial transactions and blockchain applications. There's a lot of interesting tech that's gonna be discussed today, and Albert's gonna lead that process and introduce the panel that you have in front of you.
Thanks, Larsh. Appreciate it. I'll give a little bit of background on myself before we get into the panelists. It is great to see you all here. They don't let us in technology out much, so it's a delight. Thanks very much. It's very exciting to be here on the back of RE+, right? Our industry conference was absolutely fantastic. A huge amount of excitement, which is a delight, of course, for our product and technology teams, right? We love to see the support of partners, customers, validation of everything that we've crafted. O f course, that gives us feedback to iterate and refine and further perfect our applications.
Now, in terms of why I came to Stem, it's a very interesting organization, of course. My background is, you know, technology for, you know, 25 years. I've worked at Thomson Reuters and started my platform unification activities back then. Then I went on to the CTO at Genscape. That was, of course, in the energy field, machine learning platforms, building it from the ground up and IoT monitoring, et cetera. Very exciting. Then also LiquidX, recently the CTO there. That's also about building multi-asset class trading platforms in the cloud. All of this is really around the SaaSification of applications and operating both in the cloud as well as in the edge in some instances with containerization.
Of course, here at Stem, it is all about all the above. You know, I think the other thing around the organization is a very common DNA with the essence of what architecture is really about. Microservices, fine-grained entitlements, componentization, services which can be easily adaptable to suit the need of customers and build out products almost in a synthetic way, right, through those microservices and entitlement system. The organization, you know, really impressed me in terms of the approach taken with platform, the adaptability, extensibility of it, and that was absolutely fantastic. The edge side is exciting as well. Athena Cloud, Athena Edge, all wonderful. The organization is very clearly product led.
You know, having been through a variety of situations where of course it needs to be a balance, but product led is absolutely fantastic. I see that here in spades, long-term roadmap planning is part and parcel with the DNA of the organization. That's also something I really value. The exceptional team we've seen from Bryan and others, and as you'll meet the panelists here, of course, you could probably take this slice across, you know, any number of areas within Stem and you'd find, you know, similarly qualified and talented candidates. An honor to be here. Thank you very much for listening to us. We'll get on to our roundtable now, and why don't we let the panelists do brief intros.
Hi, my name is Vikas Walia. I'm the Chief Platform Architect at Stem.
I'm Bryan Ho, Senior Director of Product Management.
Hi, I'm Pranav Aggarwal. I'm the Head of Data Science at Stem.
Good morning. My name is Sharifa Dunn. I am on the product team at AlsoEnergy.
Excellent. The objective today is to give you guys a little bit of an experience and kind of look behind the curtain, of, the magic that happens and talk to the folks who are literally designing and building, our applications around Athena, etc. To do that, we have, gathered 3 themes that we wish to talk about. Agility, simplicity and, platform intelligentsia/machine learning. I think all very kind of interesting topics as it relates to other things that you've heard today. Also with respect to agility, you hear many organizations talk about it. They say, "Oh, you know, we're agile." Right? There's very much to that which has to be carefully managed, especially as we are, 25 squads or scrum teams, you know, working to build enterprise-grade applications.
As I mentioned, architecture is absolutely essential to the go-forward strategy. That architecture needs to be evolved very carefully, even though you're releasing in production on a daily basis, which we'll talk about and how we do that. Agility is also part of the culture. It has to be very carefully managed through process, and architecture has to be established in really kind of a long-range view, and you can get out of sync if you're too active or too hyper. Agility is essential, and also architectural evolution is essential to manage carefully. We'll talk a little bit more about this. Bryan, can you talk to us a little bit about how the PD process is unique at Stem?
I think, you know, what makes Stem unique, particularly in, you know, kind of energy industry, is that it is a software tech company. From the ground up, agile processes, you know, are implemented in a true form. As a product manager, that's gives us a really sort of exciting, you know, product development organization with which to partner. You know, development's measured on the scale of weeks. You know, when we've got, you know, any sort of production issues, those are, you know, resolutions measured on a scale of hours. You know, particularly in the storage side, where, you know, the industry is still, you know, relatively nascent, right? The revenue streams, the stacks of revenue streams, the regional programs are still evolving. That agility gives us really powerful foundations by which to lead into markets.
You know, Stem's led into California, Stem's led into Hawaii, Stem's led into Ontario, Stem's led into Massachusetts. We can be on that cutting edge because we've got that machinery behind us. I think, agility is also important because we've got an open platform, which means we've got on the edge of our platform, on the periphery of our platform, these integrations with a lot of data providers as well as third-party software services. As those move and evolve, we need to be able to adapt with them. Again, you know, agile helps us get there.
What's great about that too, of course, is that as we've talked about, our team is global, right? We have, you know, an encapsulated set of scrum teams which are in India. The product owners as well as technologists are there. We all share kind of the common approach and DNA. When Bryan is working, you know, here in the U.S., right? We also have counterparts in Delhi that are picking up, and so we can operate very effectively that way. In terms of that design for integration, right? It eases kind of what we talked about earlier, is the many configurations that happen at a client site, right? Those integrations have to be handled in a natural way, right?
We kind of, you know, have this kind of configurator or, you know, the process of transmogrification of edge configurations into the platform. That also goes to how we, you know, live, eat, and breathe agile. The other way you can see, of course, is what we've done with PowerTrack, which is obviously the fantastic acquisition that Bob told you about earlier. Sharifa, can you tell us a little bit about how that was done so rapidly in terms of integrating with Athena?
It was actually quite remarkable when you think about the size and the breadth of the PowerTrack application. We have about 200,000 sites, you know, and numerous assets that sit underneath that. You think about bringing that onto the Athena platform and to have the team and the architecture to do that as quickly as we did, which was in about 6 months from the closing of our acquisition, really speaks to the right platform and the right architecture. Being able to extend the Athena platform has been crucial in bringing PowerTrack on as quickly. We think about now being able to share data across, you know, different sites, across PowerTrack, across some of the applications that we see on the legacy Athena platform that Bryan demoed earlier.
That really speaks to how quickly the team has been able to focus and connect sites and connect information for our customers, which has been super important. You know, just after the acquisition closed, we had customers asking us, "How quickly are we gonna be able to see some of our sites on the Stem Athena platform?" We were able to do that in such a short amount of time. It was important for us to be able to meet our customers where they are, especially as they're growing their sites, growing their portfolios, and being able to see across different types of applications what their assets are doing. We were really focused on bringing the team together, both on the AlsoEnergy side as well as on the Stem side, to really bring PowerTrack quickly onto the Athena platform.
It really speaks to, again, the breadth and the depth of the platform, of the team to be able to do that as quickly as we did.
I think also, you know, it's a great example of then how, you know, we can take persona-based development and do something like that single pane of glass view, and integrate then something as significant as AlsoEnergy in such a rapid time. Again, this speaks to architecture and to that point then Vikas, as Chief Architect, can you give us a little bit of an insight in terms of how you believe then our architecture supports the PD vision of Athena?
You know, like you heard, we have an ever-changing landscape. We're integrating new partners. We're running in multiple geographies, and continuously operating sites. We need to plan for change, embrace change, but we also need to keep the system up and running 24/7. Agile DevOps and our investment in, you know, continuous integration and continuous deployment, is key to that particular, you know, development philosophy. Let's delve a little bit into, you know, what we mean by that and how we practice that. You've all heard about microservices, probably not heard about micro frontends. You saw the 3 layers that Larsh talked about. You know, we've got the app layer, the Athena portal apps, and I call them micro frontends.
The Athena Cloud services, those are microservices, and the Athena Edge services as well. You know, those are services and components as well. I call these micro components. The key thing that we care about for these micro components, you know, for scalability and reliability and availability, is for us to be able to independently deploy them and independently scale them, right? That's the mandate, right? Because you have different demands for different components. Those are the 2 key aspects, if anything, that we care about. Now, where we take from there is we've got, like, 80+, you know, micro components identified, right? We've got different guild members. We've got people who are UI specialists, we have people who are data scientists, people who specialize in data pipelines, and they've got their own ownership of components.
We take those components, and we assign them team ownership. Then to those team members, you know, they're responsible not just for development, but also for test automation and for, you know, deploying it in production and also operating it, you know, and monitoring it and feedback into the loop, right? It's that what we call agile DevOps, right? The teams who are responsible for that component are responsible for the entire lifecycle and upkeep for that particular stuff. The next thing that we enable them, so our CI/CD team, we call them developer experience team, and what their focus is on empowering these teams to be able to iterate quicker, right? It's not just sufficient to-
You know, we can no longer have the luxury to say, "You know, I'll deploy this when I have all the pieces of the puzzle figured out 6 months later, 9 months later." You know, what we wanna be able to do is iterate quicker, you know, deploy quicker, make small changes in the system incrementally, and make that change and let the developers embrace that change on a daily basis. What we've set ourselves up for is to take any of those components that you saw, whether it's the micro frontends, the microservices in the cloud, or microservices on the edge, for developers to be able to develop them every day and push that change out every day. It requires a consistent effort and a consistent practice and a discipline and rigor in the organization.
You know, to make sure that your code is being executed, it's being built every day, it's being tested every day, it's being deployed every day, right? We have around 80+ components across the edge cloud and micro frontends. We run on average about 3 component integration test cycles, where we actually deploy that software with all its dependencies and run it through its paces every day. We run on average about 5 to 10 system integration test cycles, where we deploy the entire suite of apps and run it through its paces. Those are components which are, again, you know, because of the practices that we follow, we can simulate the entire deployment pipeline, run through paces, and then allow our developers to deploy at will.
To add to that and complement it, I think, you know, when people talk about agility, then they talk about continuous deployment, they talk about being able to make these changes in rapid cycles, right? The responsibility is huge on the development of the QA teams, right? Because we're powering the grid. E verybody has a very deep sense of quality to that point. We've invested heavily in QA automation. T hen also in terms of how this relates to agility and functional changes, right? On the front end, right? The customer's experience, right? It's also on the back end, on underlying logic, and also in terms of models and algorithms, right?
Being able to make changes and enhancements there to models and features is important. To that point, Pranav, can you tell us a little bit about how our operating rhythm supports the AI brain and machine learning that you run?
A bsolutely. As head of data science, my team's mandate is to release cutting-edge algorithms and intelligence into the Athena platform. Being able to do that at scale, at speed, right. In an agile manner, in an ever-changing landscape. We have hardware complexity that's changing. We have tariffs that's changing. The market constructs are changing. In all of this landscape, we really need an agile process to be able to iterate and sort of release cutting-edge features into the Athena platform. We leverage what Vikas and his team has built in a big way to be able to not just deploy software in the cloud, but also on the edge, which is extremely hard, and being able to do that consistently and reliably is super important.
Another thing that we do extremely well at Stem is once we've released the product into the market, we define performance KPIs that are continuously being fed back to the engineering team from the fleet, from operations. I run a weekly call with performance and operations team, where the performance KPIs from the fleet are being fed back, and we're learning from the fleet.
No matter how much we simulate the hardware behavior in a big sense, we learn from the operation in the markets, right? So as we get these learnings, we need to iterate and sort of keep our product performance on the cutting edge. Having that agile way and mechanism to deploy the software in cloud and edge is an extremely big enabler for us to sort of deliver good quality product consistently.
To that point, I think, you know, we can get very excited about our technology, right? Containerization and cloud containerization at the edge. That's a very refined workflow and deployment model. Doing it within Athena is fantastic. While we're excited about the technology, you know, it has to correlate to the financials, right? What's the so what, right, for customers. Bryan, can you give us a few tidbits and factoids on the financial benefits?
I mean, you know, like Albert, you're saying, I mean, all of this, right, I mean, has to come at the service of a customer, at the service of the asset owner. I think what this agility does in addition to, you know, some more thoughtfully planned development is it allows us to, you know, be opportunistic and solve hard problems that other people can't solve. 1 example, you know, is there was an asset owner who had a, you know, fleet of 86 sites in California that were underperforming, and they wanted to bring on new software to improve the performance of those sites. Taking over 86 operational sites, you know, in itself is a, you know, feat that, you know, very few operators, you know, very few operators have 86 sites, right?
Taking over a new portfolio of that magnitude is, you know, an undertaking that I think probably, you know, Stem is the only, if not one of the few people who could pull it off. We brought those sites over into Athena, got them operational in under 60 days, and then within the Q1 of operations, demonstrated a 30% improvement on performance on those assets. That's. Those, like, that's the kind of, like, magnitude of problem that Stem can solve with this agility that's given by the product development team.
Perfect. W e have kind of a, you know, a myriad of these case studies which are, I think, very interesting and illustrative of what has gone into our platform. To that point, simplicity, right? Simplicity is another theme as we move on to the second one. How do we delight customers? That customer satisfaction comes in many ways, shapes, forms. For example, it can be something like the 86 sites and how we absorb those, which is much around the integration, but it's also about how we integrate applications, whether they be third party or acquisition, you know, and coming and building those applications and presenting them through a single pane of glass. Right? You've heard a bit about that. We've heard a bit about PowerTrack.
You know, from the simplicity perspective, there's a lot of engineering that happens under the scenes or behind the scenes, behind that curtain that we're pulling back. You know, and there's a lot of blood, sweat, and tears that goes into that. A t the end of the day, customers experience it through that single pane of glass. With that, Sharifa, can you tell us about how then our Athena platform and our process there of simplification as you use the PowerTrack integration?
A bsolutely. When you think about PowerTrack now being a part of the Athena platform, we are really now one of the few, if not the only provider that provides asset performance management and economic optimization under 1 roof. That really translates to simplicity for our customers. We are positioning ourselves as a solutions company. The markets are nascent, as a lot of my co-panelists have mentioned, and we need to be able to be flexible, to be modular so that we can meet our customers where they are. That means a lot of different things for, a lot of different people. When we think about the stack of solutions that we provide, we can do everything from the edge to the cloud, and therein.
We can provide modular solutions to our customers depending on if they need just solar, if they need just storage, if they need solar plus storage. We have applications that are set up so that they can maybe buy one of our applications or buy multiple of our applications. That's really the key to the success of our platform here and why we're setting ourselves up to be simple and flexible so that our customers are getting exactly what they need, exactly when they need it. That allows us to continue to layer on applications, as well as features and functionality, so that when our customers continue to grow, add assets, add markets, become more global, we are able to grow with them. We've heard that theme before by a couple of speakers ahead of this panel.
We really are making sure that flexibility is a part of the ethos of our teams, and making sure that we are able to appropriately and sufficiently meet our customers and meet the markets, especially as things are changing so quickly and so dynamically. We're setting ourselves up really to make sure that our applications and our platform are really set up for that success. Really another key point around that is building applications that also meet those personas. You heard Larsh in his section talking about, you know, PowerTrack being for O&M providers, Analyzer for developers, and making sure that we're crafting applications that are really suited for specific personas so they're, again, getting exactly what they need when they need it.
when we continue to integrate and extend our platform, thinking about OEM integrations, thinking about utility interconnections, being able to partner with folks like InCharge with Sitetracker so that our customers are able to plug and play into our Athena applications, into our Athena platform where it makes sense for them. That's really a call to, again, flexibility, which provides simplicity for our customers.
Thank you. T hat flexibility correlates to the modularity of the platform. Modularity with services, microservices, and again, back to entitlements. To this point about modularity, you know, Vikas, can you talk a bit more about the architecture and the simplicity that we build in, part of that also being trust?
I'm gonna be the developer evangelist here because, you know, I firmly believe that if we simplify the life of our developers, we empower them to delight our customers with, you know, more powerful apps, and iterate more quickly. The choices that we make and, you know, how we do things is motivated by that, right? How do we simplify the life of developers? One of the conscious choices that we made, you know, a long time ago, 5 years ago, is we were early adopters of containerization. Ever since, like, 2017, we had, like, 100% deployment of our production using containers, right? It simplifies the life of developers, whether you're a Python developer or a Java developer or, you know, data scientist or UI developer.
You know, you have different development stacks, but it homogenizes how you develop and how you interoperate in that particular system. When we set up the environments, whether we're using. You know, we use a lot of simulators, you know, we use device simulators. We have our own software that's deployed on the edge, and we go a step further. To day it's become quite a bit of a fashion that most of the cloud deployment is containerized because of the, you know, immense benefits with that. 5 years ago, it was not so, and we made early bets on that, which, you know, which proved out to be quite rewarding. We went a step further and containerized and homogenized not just on the cloud, but on the edge too.
Even today, you have, you know, very few companies who roll out the software on the edge. You know, edge is the new frontier, right? It's an important piece of our software that runs on the edge. You know, the real-time controls happen on the edge, and we have to iterate software for our controls on the edge with the same agility as we do in the cloud. Having a homogeneous way of deploying for developers really eases the friction in terms of how a developer develops on their workstation to, you know, deploy it in the cloud with simulators to push it out there, whether it's the edge or the cloud. Another example of nuanced process but simplified is our approach to cybersecurity, right?
We are operating critical assets, critical infrastructure, you know, so we need to do it reliably, and we need to do it securely. All of us know that, you know, cybersecurity threats are pretty real. We invested in a nuanced approach within the organization, engaging, you know, the people, process, and technology in a fairly simplified way, you know. Let's look at what we do in terms of, you know, some of the things. It starts with training people not in, you know, long training sessions, you know, once a year or when they start.
Our approach to training people and educating them and reinforcing that learning is basically training them in bite-size, you know, 5-minute, 10-minute, you know, courses that we send it to them, which are educational, and they're also testing their knowledge of cybersecurity, right? And that's something that applies to everybody in the organization, right? Because somebody in the organization clicking an email with some vulnerabilities is as detrimental to, you know, a developer doing the wrong thing from a code perspective. On the process side, we focused on SOC2 compliance. You know, we passed with flying colors in record time from start to finish. On the technology side, you know, we embraced that continuous development and continuous patching and continuous software upgrade all the way in terms of the CVE scanning as well.
We set ourselves on a process, again, on a leading edge. Instead of like you probably heard about people scanning or, you know, scanning the SAST and the DAST tools out there. We're early adopters of container scanning. When you package your software in the container, you've got different components. You've got components that are written in C ++ , you've got components written in Python, you've components written in Java. We wanna scan it all together and make sure we're addressing the vulnerabilities all the way from the Linux operating system, all the way to components that we develop, right? There is a software BOM, and we need to continuously scan our software BOM to see if there are vulnerabilities out there that have been found by, you know, researchers out there on the internet.
We scan all of our containers every day. We find vulnerabilities every day, and we patch them every day. We've incorporated that into the same agile development process. To date, this year alone, you know, we've patched about 3,000 of these vulnerabilities in our software, right? That's a scale of engagement of our developers into that particular security-minded process. As much as, you know, it is a nuanced approach, you know, we invested pretty heavily in developer experience to make it seamless for developers to work and make it part of their daily rituals, right? It's not something that they do as a one-off.
They follow the exact same CI/CD processes, exact same deployment, and we try to do as much of these things ahead of time for them to the extent that for most of those changes, all they need to do is accept the merge request, accept the patch request. Our investments in packaging the code, our investments in testability, in testing the code on a regular basis, you know, does the bulk of the work for them.
I think, as you hear Vikas talk about that, you know, this really does place us, as you saw in John's presentation, at the top 5% to 10% really of digital leaders. Being able to containerize in cloud and on the edge and deploy in real-time, continuously run InfoSec, and ensure that everybody has that same mindset, is trained in the same way, as we accelerate with our growth in India, et cetera, our development team there. It's relatively unique. I n addition to that uniqueness then, you know, is kind of the AI and ML and optimization that we do with forecasting and all of our modeling, as you've seen in the applications. You know, it looks relatively simple. Underneath that, there's a lot more to it.
Pranav, can you give us a little view about the wizardry that you do?
Yeah.
for optimization?
Absolutely. I mean, I would really like to underline the point of having a microservice-based architecture that we have. It's really important because we have structured the optimization and the core forecasting functionality as independent microservices. What that lets us do in a big part is if you view the customer engagement with our product through the life cycle, when the customer comes to us and they want to assess the expected performance on the site, we are using the exact same microservice of optimization. Which is the same thing that we use when the asset is actually in, on the ground and operating. We're using exact same microservice. Then later on when the asset performs, right, and we want to kind of compare the performance with the expected performance, we're again using the same microservice of optimization.
What that really lets us do is it holds us accountable and honest to our customers of what we are promising early on during the pre-performance stage is exactly what we deliver them during operation. That, that's a huge enabler in being able to leverage microservice-based architecture to keep things simple for the customer, whereas all the complexity is sort of containerized within a single microservice. Very recently we saw, like this year in California, we have dispatched 72% year-to-date higher number of grid dispatches than what we did in 2021. From a customer's point of view, we are experiencing higher heat days in California, and the system is performing. It's dispatching when the utility is asking it to dispatch.
It's, it seems fairly simple, but in the back end, the all of these microservices are orchestrating seamlessly to kind of ingest the OpenADR signal, run this forecasting, run optimization microservice, and dispatch the asset. All of this complexity is kind of under the hood. From the customer, it's a seamless experience of delivering the performance when it is demanded from the utility, from the grid.
Thanks, Pranav. Carrying on with that customer centricity, from a product perspective, Bryan, can you give a view on how simplicity is enabled by Athena?
Yeah. I think in terms of, you know, customer-centric value derived from, you know, simplicity, I think. Well, really, you know, we, to a degree, we do put on kind of a magic show, right? I mean, like, Vikas and Pranav said, there's incredible amount of complexity that goes on under the hood. T o our customers, as you see through the apps, through the lens that they see things, you know, these are pretty charts and these are, you know, performance numbers that should be hitting the mark. I think, you know, we put on a happy face to our customers and create a, you know, a simple experience for them, right?
To ensure that, you know, we are doing our jobs on the product side, on the engineering side, you know, to ensure reliability of performance, reliability of operations, so that we do not suck our customers into those issues, and we can ensure that what they're worried about is developing projects, right? The getting projects from the ground, getting projects built, managing the high-level operations, you know, of their asset fleets versus having to get into the weeds with us. That simplicity is hard. In concept, right, I mean, you know, you can think of, again, forecasting, contract management, optimization, and dispatch as a pretty simple fundamental loop. Under all that is, you know, the operating history that Stem has, and I think the number is somewhere north of 25 million hours of operating runtime now, right?
Has kind of given Stem all the bruises and taught us all the lessons we need to understand what you need to reliably operate 24/7 in a fault-tolerant way. You know, data connections go out, meters go bad, you know, software services fail. Even AWS, you know, can have service issues. H ow do you operate through all of that, and maintain performance at a high level and ensure again that all of that is abstracted away from the customer experience? There's been an immense amount of investment done at Stem over the years, as a leader and a, sort of a longtime player in the industry. A lot of complexity that's been built into the back end to make sure that, you know, we can operate reliably.
Great. Thanks, Bryan. It moves us on to our third and final theme, which is around platform intelligence and machine learning/AI. As you guys have heard, a lot of the secret sauce in the platform is around that, right? Devising the analytics around forecasting optimization, you know, discharge events, et cetera, right? That is a very complex task, and the Athena platform does it remarkably well. In order to do that, we've talked about how ingestion of data is essential, right? We've heard 1 TB per day. We're monitoring over 500,000 endpoints, which is fantastic. To feed the platform and that intelligence around it, you know, we need to empower the platform with AI and machine learning.
Vikas, can you give us a perspective on how you have crafted Athena, with AI ML?
Yeah. We really kept it simple. For our AI ML developers, you know, we empower them in 3 different ways, right? The first is with big data, and making that big data readily available to them, you know, from production in real time.
The second is big compute. You know, we have AI ML processes that are, you know, quite hungry for compute, and I'll talk about that a little bit later. The third is hands-on engagement in operations, right? Let me focus on the first one first. We get about a terabyte of data from our sites every day, you know. That's a lot of data to crunch in real time. We invested in using scalable technologies only possible in the cloud, right? With that kind of volume. We make all of that data available and the curated insights on top of that to our data scientists in near real time. They can pull the data at any time for their simulation offline or online without affecting the production system for the life of the system, right?
We've got 10 years of data for some of the systems. We can pull that out in a heartbeat, right. Second aspect is, you know, compute. As you all are aware, you know, AI ML, there are certain aspects of AI ML that are very compute intense. T here are certain aspects in a way that we use it, we want that to be not a barrier. We create models, unique models for every one of our sites. We don't have a common model for all of our sites. We create a unique, you know, forecasting model for each of the, you know, things that we forecast for each of the sites. T hat requires model selection, and that requires model training.
We, for every one of our sites, you know, we do model selection, which is triggered either manually or automatically. When we detect deviation of the forecasting, you know, algorithms, then there could be model selection that's triggered. Again, you know, that thing could be triggered for all of our sites at the same time, and that could result in spinning up on demand, you know, hundreds of containers, and we have. That's effortless scalability that we offer to our AI ML developers, saying, "You got big compute at your disposal, use it." We do model selection every month, we do model training every hour for every site. You can just imagine what that volume of, you know, processing that goes on. The last thing, you know, which is hands-on engagement with ops, right?
When things go wrong, and things will go wrong, you know, they don't behave, the assets are not behaving per your expectations. That could be because of human error, that could be because of software errors that are found, or that could be because of conditions that we didn't anticipate in our design.
The good thing is that our data scientists, we don't have them locked in an ivory tower, fed some reports, you know, once in 3 months, and then saying, "Can you come up with this new algorithm for me that fixes this particular problem?" Our operations feeds in through automatic and manual means, shoulder taps our data scientists right away, and they jump on the system right away to see the behavior of the system as it's as the anomaly's unfolding, and contributes directly in terms of either interceptions that can be done. Sometimes we have a 4-hour event.
You know, we see something going wrong in the first hour, and sometimes we engage, you know, folks in the first hour, and they can recommend certain interventions, you know, certain knobs and dials that they can play with to basically make sure that we have an overall, you know, better outcome in that 4 hours. That's the agility that we offer to our developers. Again, when things go bad, you know, we also allow them to be able to make remediations in an agile manner. You know, push code out, develop remediations, test them quickly, and push them out in less than a day.
Thank you, Vikas. You know, without getting into further trade secrets, et cetera, I wanted to wrap up our roundtable. Thank you all for the time and listening. As you can hear, we're product-led, tech-enabled, and customer-focused. Thank you very much. Over to Larsh.
Yeah, I love these guys, and you can see why. These guys are basically, you know, the best in the business. They've been with us for quite some time and working on all these different things that you've heard about. I'm assuming that at this point, we can all give you a software test, and you'll pass with flying colors and submit your resumes for software developer positions 'cause it's a fascinating place to work, and I'm sure you would all enjoy it. I'm gonna wrap up really quickly 'cause I know we need to get to Bill. I just wanna give you a little bit of that future of Athena.
You can see in the image here that we are creating this vision of multiple applications, the modularity that Sharifa mentioned, the kind of capability and the flexibility for, you know, persona-based development to really simplify the operation of clean energy assets and add to the capabilities of the system a number of different dimensions, whether it's new use cases, new geos, you know, new equipment types, new assets of all different models, and then certainly adapting to the continuous transformation, you know, as we accelerate this energy transition. With that, I'm going to hand it over to Bill Bush.
I appreciate everybody coming today. This being our first official Analyst Day, we thought we would run through some of the highlights of the business. Talk about target metrics, which I'm sure probably most people have already downloaded from the web based on the number of computers. That and then talk about some of the potential risks that the business can run into, some of which we've talked about at some length, and others that are new, which also are opportunities for the company. Since we've been public, I think the company has really performed well. You can see we talked about the revenue CAGRs already, the positive GAAP gross margin.
I think one of the things that we're most proud of is the growth in CARR, contracted annual recurring revenue, almost 110% year to date. That's as you think about the company, that's definitely something that you'll want to focus on. I mean, that is the lifeblood of the company. It drives the highest gross margin products that we have, and it really differentiates us from others in the marketplace. Operating leverage improvement, which certainly in no small part, thanks to the Gibson Dunn team. We are using their space today, and that's obviously part of our operating cost leverage here. I think one of the things that I think we've been very fortunate about is the capital side of the business. We had a very successful IPO back in April 2021.
Not so long after that, we were able to extinguish nearly all of the warrants which were outstanding at that point in time. We've been able to both combine a green bond under very favorable terms with an acquisition. I think all of those have really been strong aspects of what we've done from a capital standpoint. You should expect to see those sorts of things from us in the future. When we got started back in April, we had 2 covering analysts, and we now have 10. We also have strong average daily trading value. I think all of those reflect both the reception of the company, the business, and the work that we've done in terms of making sure that we're both public ready and public, you know, into the future.
One of the things that I wanna make sure that we kind of step back a little bit is like what were some of the thoughts of the senior leadership team as we went into the going public transaction and since then? These 3 things, growth, profitability, and predictability. That's what we've been talking about within the company for years now, when we both were private and invested from the private equity side to today. I think those are the types of things that you should expect to see from us in the future. That's how we feel very confident, both in our guidance, which we'll talk about here in a couple of slides, and the statement that we think that we're gonna be EBITDA positive in the H2 of 2023.
We have great confidence in that because of the predictability component of the business. Our FP&A team works seamlessly with the various product and supply chain teams and, you know, across the business, and so we're really able to put a fine line on what we're doing. One of the things that's been super helpful to us is that the AlsoEnergy team, just as a for instance, also use NetSuite, which was certainly a part of the decision to acquire that company because we use NetSuite as well. It allowed us to bring them onto our platform and file our first 10Q with them, you know, within 45 days from when we had completed that acquisition.
There's a lot of work that goes into things like that, but that's where that predictability comes from, which is systematic processes to be able to create financial records as quickly as possible. The 2025 financial targets. I'm sure most of you have already seen this, one of the things that this is the result of a long-range planning process that was led by Jigar Shah, who runs FP&A for us, and all of those teams within the company. One of the things you'll probably see that comes out on the page is that one of the things that we expect really are 2 pieces. The first is that software and services is gonna grow somewhere between 65% to 85%. We feel quite confident in that. Again, highest margin, highest gross margin part of the business.
The hardware, though, we expect actually grow slightly slower than the marketplace. The reason for that is a good reason, and I think it'll inure to us positively, is that we expect to do more software-only deals going forward. The hardware component of the business probably will slow down some. We've been talking about that in the public marketplace for a while. As we move up the size, scale of the business, we likely are not the best supplier of hardware. A good example of that is the Available Power deal. Size deals like that, we're actually supplying hardware to that. As you go forward, the balance sheet that we have in the company isn't necessarily either gonna be appropriate, nor will what we would describe as a fully enabled developer look to us to supply hardware.
What they will do, though, and we feel very confident in that, is allow us to continue to provide design-based services to those same systems. One of the things that you'll see less is the sale of hardware. One of the things that you'll see more is services revenue from the business. In the past, we really kind of bundled those 2 things together. We're now, as of August 1st, separating those items. You'll see a growth in service revenue in the total business going forward. We're really happy about that. Of course, as we've said, you know, the hardware is the least differentiated part of our business. You know, you just heard from the tech team, that's where the differentiation is. That's why customers come to us. That's why we're continuing to win in the marketplace.
The non-GAAP gross margin range of 25% to 30%, really reflecting all of the work that the company is doing in terms of managing its growth and focusing on an ROI-based methodology in terms of deciding which projects to work on and which projects to fund. Of course, lastly, adjusted EBITDA margin. As I mentioned, we expect to be EBITDA positive in the H2 of 2023. By the end of 2025, we would expect to be in the range of 15% to 20% on the adjusted EBITDA level, which reflects a lot of the work that we're gonna do, both on the top side and at the operating leverage line, which we'll talk about here in a second. Operating leverage, and I think this is a big part of what we're doing.
Really, one of the things that I think has been the most successful part of the business has been our ability to drive operating leverage. One of the things that came with the Also Energy team, and we'll talk about here in a second, is a very strong international operation in India, which is going to allow us to continue to drive down our average wage per employee. One of the things that we look at and cross this spectrum is really the ability for us to drive more sales, drive more gross margin at an ever-decreasing percentage. We've been able to do that so far. We think that in the future, we're gonna continue to be able to do that at an accelerating rate.
Talking about India, I think this is something which I think is aspirational for a lot of companies. They talk about their ability to diversify their workforce. The good news for us is we've already done that. Through the AlsoEnergy team, there's 100 folks in India already. We expect to increase that number significantly. If you think about the hard work that Bob and his team put in, that's already a part of our DNA today. The ability to add that marginal person in an overseas environment is significantly easier when you're at 101 to 102 versus 1 to 2. I think we're set up for success to be able to do that.
Across the platform, we're already moving significant amounts of our transactional work to the office in Delhi, and we expect to continue to do that. That sort of task and opportunity for us is gonna be the thing that will drive operating leverage for us. We're really excited about that as an initiative. Risks that we're managing, and I think this is a significant part of the planning that goes in or went into the LRP. It's gonna be a significant part of what we're focused on in the future. Some new things, I mean, the IRA is obviously a huge boon to the business, but there's also issues that come with that.
You know, when you think about what's gonna happen in terms of supply chain, what's gonna happen in terms of how the transferability of tax credits is gonna work, much of that is not yet quite known. We think that there's a huge opportunity, obviously, but there's work that has to be done there, and so we'll continue to do that work. UFLPA hasn't hit the battery side of our business yet, and I say yet because we don't know that it will or won't, but certainly has been on the utility scale side of the solar business. We've seen delays in terms of transfer of panels into the United States. Whether that increases in the future, we're not sure, but it's something that we're closely monitoring for impacts on our business.
The next 2 are really ones that we've been talking about for some time. Supply chain constraints. One of the great news of the, you know, the IRA is the storage ITC. The potential downfall to that is that there's not new supply coming on in the next year that, you know, matches up with the potential demand. We've long been talking about oversupply in 2023, in late 2023. We're not sure that that's gonna continue to happen. Good news for us is that we, as John mentioned, we have supplied all of the equipment that we would think we would need through the Q3 of 2023 and a significant amount both for the Q4 in 2023 and in the Q1 and Q2 of 2024. We're in good shape from that standpoint.
More to do, but it's for sure we're in a very good spot from that standpoint. Then, of course, you know, interconnection and permitting continues to plague the industry. I think Senator Manchin is talking about having a bill. We're not sure what the impacts of that bill will be, but most of that work is actually at a local level rather than federal. It's something that I think has impacted the business negatively. We've talked about that some in the past, and I don't see a near-term solution to that. We'll continue to work with the interconnection and the permitting agencies that are part of our network.
Much like on the solar side, which I spent more than 10 years in, you know, there were ways to make that process more seamless, and we'll certainly work with those kind of parties to do that. As I mentioned, 2022 guidance, we're reaffirming that. We don't see any risk to where we are at this point in time in terms of that guidance. Business is growing well, demand's very high, and so we're excited about where we are from a 2022 standpoint. What you'll see from us in the future, just to, you know, kind of level set everyone, we will be doing this Analyst Day, providing longer range guidance. In the February timeframe when we announce our annual results, we'll give the next year's guidance.
You should expect from us 2023 guidance in the February timeframe. We haven't scheduled our earnings call yet, but it will happen late February or mid to late February. We're in good shape for 2022, and I think the business is well positioned for the future. A couple of quick things just as you're thinking about next steps in terms of the business. What you should expect from us is that we're gonna continue to focus on TAM and the most profitable aspects of that TAM as well. When you think about, you know, business that we're going after, we're not just chasing deals. We have an ROI focus. The sales team is well integrated into this. The sales commission plan, importantly, is also focused on that.
There's a lot of positives in terms of what we're doing. All of that's gonna result in increasing margins and profitability of the business. We're very focused on being a capital-light business. You won't see us going and buying a lot of assets or anything like that. We're gonna continue on this focus. We're a software-focused business. What that will mean is that EBITDA and free cash flow are effectively the same, not significant differences between those. We do have some amortization, depreciation, but not significant. Really, EBITDA will be a good proxy for free cash flow. Again, EBITDA positive in 2024 in the H2 and targeting 15% to 20% out into 2025. I think we're in great shape as a business. We've got the team in place.
We continue to add strong folks across the business, and we're really looking forward to the future. Those are the financial slides. I know we're gonna go to Q&A as well. There is lunch out on the other side. I think maybe what we were thinking to do is everybody wanted to.
Let's go do a little Q&A here before we do lunch.
Perfect.
If we can have all the team come up. We'll be over a little bit over 12, but we'll maybe start with questions from the audience, and then we've got questions that are coming from the web. If you're on the web and you wanna ask questions, go ahead and use the chat function. We'll start in the room here. Grace.
Goldman Sachs, this is Grace. Appreciate all the color on IRA and the potential upside. Just quick question on the $6 billion retrofit opportunity. It seems like could be a big upside, but that seems to be gonna be more like in the back half of 2023. Is that fair? Just wonder how quickly can you leverage that $6 billion. Let's say, what % of the $6 billion can you get by 2025? What is the reasonable share you can achieve?
Yeah. You know, one of the key things I wanted to emphasize in that discussion was us building the tools to automate attacking that market opportunity because we think it's more expansive than just within the AE footprint. If you look at the broader C&I marketplace, our Fortune 500 customers or the asset managers that we have relationships with, if we can target sales to them in an automated self-service way, here's your pro forma economics. Push here, you can expect delivery of the storage system by this date, and it's backed by partners of ours that can provide the financing.
We think that is a way to reduce customer acquisition costs, drive operating leverage, and really serve a segment of the market that's very difficult for other competitors to access because it's fragmented, and it involves a lot of hand-to-hand combat if you don't have the tools, the data, and the relationships to execute on that. We'd like to get that right. We think we're very well-positioned, certainly on the AlsoEnergy footprint to do it because we have, you know, upwards of 40,000 customers that are happy. We have a contractual relationship. You know, that $6 billion just in the markets where we're transacting today feels very achievable. That's 30% of the total footprint. That feels very achievable. We haven't really broken out, you know, expected win rates. F rankly, you know, it's early days for that.
We expect revenue to lag bookings. We are expecting bookings in 2023, but the revenue will really come 2024 and beyond.
Hmm.
Thanks, Joe. It's from Guggenheim. 2 questions. First one, simple one. You gave us the hardware and software revenue CAGRs. We can do the math, but if you could just clarify for us what you expect the overall revenue CAGR to be between 22 and 25? The second question is other asset classes. You've got wind, geothermal, you know, non-lithium ion storage. I'm wondering how you think about potentially incorporating those in your platform over time. Thank you.
Want to go backwards. I'll take the first part, Joe. I think from the perspective of Athena, we could put Athena on top of really any battery technology, and we do get called on a variety of those. We'll continue to monitor where it makes sense, and we have in the past, quite frankly, looked at some other technologies other than just lithium ion, as an example. Bill, you want to take the other one? We don't have 2023 in the number yet, so guidance.
Mid-30's.
Thank you.
Okay. Well done, Joe. Congratulations.
Yeah, for sure. Maybe the answer, the other part of that answer is also it's gonna be high margin focus. Mid-30s% growth, but this is really a business which we're positioning for profitability as opposed to just straight growth. I think that's really the key part of the answer is the absolute part of, you know, the absolute inflection and what the scope of the curve is less important than the profitability of the deals that make up that. Hence, the focus on software and services, which are much higher gross margin for the business.
I guess just maybe more on the hardware/software side. It seems to me that obviously with the IRA, the hardware or the market demand will be really strong. It's natural to think that obviously you have a really good software offering, but it's much easier to sell more hardware so your software can tag along, right? I just wonder, I guess I understand the rationale behind the hardware being slower, but why don't you think the other way in terms of you sell as much hardware as possible so the software growth will be much easier?
Yeah. I think the way to differentiate that is from market and/or system size standpoint. We're gonna. Nobody should hear from either this answer or any prior comment that we're not gonna be selling hardware. That's not what we mean. What we mean is that projects. Because we're significantly moving up the size curve in terms of projects, which means that what we're saying is that we're less likely to sell hardware into those types of projects. We're not unlikely to sell hardware into smaller, the traditional part of the business, which was BTM. You know, i.e., system sizes of 5 MWh and smaller. You know, so different. Those are gonna be strong margin systems. They always have been for the business. On the top end, you see 2 things.
1, fully enabled developers, meaning that they have procurement departments. 2, very strong, you know, say, a lack of differentiation in what you're purchasing. It's not set up well for us. Large capital constraints. I mean, if you think about, you know, multi-gigawatt hour systems, the amount of hardware that you're talking about in terms of dollars is huge.
That's not necessarily something that Stem, with its current balance sheet, can participate in, right? We don't wanna leverage ourselves into 1 project at the expense of all the other high-margin projects which we would have to pass on to do that. That's what we mean. No hardware, more systematic, more ROI-focused investments in hardware selling.
Ron, I'd add, you're gonna see more. I think we're seeing more incoming on the software-only piece. To take Bill's point a little further is, if they don't or if we don't wanna participate, quite frankly, in the hardware piece, we can always go with the Athena side. These large developers don't have a software solution, so they'll still need that component, which is great high margin business for us.
Biju Perincheril, Charles Schwab. Just expanding on that hardware-software split. Are you looking at potential partnerships with large integrators? You bring in software capabilities or maybe just talk about how that split happens.
You mean actually putting Athena at the manufacturer level that would then be sold more broadly?
Right. Like
I mean, Biju, certainly longer term would be our goal to see that occur. I think that it's something that we've talked to some of the suppliers about. I mean, there are a lot of discussions on hardware, software only, and the fact is our customers demand that we provide the full solution at C&I for sure. Larger developers, as we just said, will maybe do their own sourcing, but certainly use Athena. I think longer term, that would be you know, admirable opportunity for us that we'd like to see. You know, we're kind of. Right now I think we have a very nice diversification of hardware suppliers. You know, we'll keep working that as opposed to aligning with 1.
If we aligned with Athena on 1 hardware supplier, probably pretty likely that they would say, "You're gonna buy 100% of your hardware from us." I don't think that's a spot we wanna be in. It's served us well to have multiple suppliers, particularly in times like this, quite frankly, that have been a little bit more turbulent and enable us to flex with other suppliers that might have capacity.
Maybe 1, a second question was on the market participation revenue. When does, from the existing assets, when does that inflect? With the ITC change, are you now able to realize that in a shorter timeframe for newer projects?
I'll take that, Bill.
Sure. I think it's a kind of a dual style answer there. The first part is that in general, when you think about the existing fleet, those are solar plus storage, i.e., there was an ITC incentive attached to them under the old rules, which means that there are charge limits associated with the recharge of the battery from solar. For the most part, you're not gonna see those asset owners be aggressive until the expiration of that recapture, i.e., 5 years. Just in general, that's 1 way to think about market participation as it relates to solar, you know, to the existing fleet.
Going forward, I think that's a very different answer, and this is really where the work that Bob's team has done will be, I believe, very helpful to the business in general. There's 41,000 sites that the AlsoEnergy team has, our current customers paying the business for high-value services, though some percentage of those are going to need and/or want a battery attachment. Those systems will not have that constraint that I talked about, as will other systems that we can potentially sell into. I think that market participation component will be different for the future components of the portfolio, very different than the existing.
Hi, Abby from Northland. Just curious.
Yeah, the latter. I mean, we're hardware agnostic. Actually, I prefer neutral. I think agnostic means we don't believe. I think we're neutral, right? That is to say there will be asset types and storage products that are characterized by behaviors from the OEM that may limit their performance in certain characteristics or be advantageous in others. As we select the OEMs and we work with them, we understand what those characteristics may be. We've talked about other like long-duration storage. There are characteristics of flow batteries that are different that you have to recognize if you're gonna build a use case around them. We don't necessarily have a preference for an OEM or a particular product, but we do look at the product make and model.
We do evaluate their characteristics and say, is it suitable to achieve the goals for that particular site? In many cases, lithium-ion is somewhat standard and commoditized and so forth, but that's not 100% true across the portfolio of OEMs that we work with or the products that they offer. We do make that evaluation. Otherwise, we are just looking for the best price, best performance and performance guarantees that suit the duration of the project life.
Take a question that came in from the web. Talk about battery supply and labor. Are there adequate crews to install all the battery banks?
Take it, Bill?
Sure. We of course are not a contractor, and so it's important to say that that's not part of our business. We are obviously impacted by a number of items, and certainly crews are one of them. At this point, though, we don't see labor as being a blocker in terms of advancing projects. You know, and that's really more from our partners on the EPC side. We just don't see that as an issue. Permitting and interconnection are much more problematic than labor. That's not to say that if you solve 1, you know, it's the Pareto. Could labor eventually become a problem? Potentially, but it's not the biggest problem today.
Thank you. How do you think about when you put up the 2025 plan, the sort of, mixed component, U.S. international relative to today? How do you see that sort of developing over time? Thank you.
Thanks for that question. When we went through the LRP, I would say it's almost a kind of pre-IRA and post-IRA view of the world. I think pre-IRA, we thought that the international aspects were gonna be much more significant to us, really building upon the work that Bob and his team have done. Post-IRA, I think one of the comments within the company is that maybe the best international market is here in the U.S. You know, so I think as the IRA gets more clarity and there's a couple of important provisions, you know, the transferability of credits is 1 that I mentioned. There are others as well that need to be clarified by the IRS.
You know, I think that for us right now, I would say in the near term, we're gonna really take advantage of what's happening here in the U.S. I mean, the good news is that we have assets, i.e., teams internationally. We're doing work in Chile and Colombia that Prakesh has led for some time. We're gonna continue to look at those. A gain, I'd come back to an ROI focus in terms of the business driving margin, driving profitability. In general, you'd say, like, "Hey, going to, you know, fill in the blank foreign country versus taking advantage of an opportunity in Arkansas. I'll do the ones in Arkansas.
Yeah. I think the real question too becomes, is there an IRA-like initiative in another country? If that's the case, I think we'll be very well-positioned through Also Energy in the sense that we have a hub in Berlin that we could address any European action around those lines. Then there's a group in Japan, so that we could spread around Asia. I mean, to Bill's point, we feel like it's such a great opportunity here that we're gonna probably stay close for a little while.
Couple questions. Craig at Tuohy Brothers. First, on slide 69, I noticed the bidder and price manager applications coming soon. Some of your peers have made a big deal about highlighting the balkanization of the markets and the different bidding and revenue opportunities from market balancing to ancillary services, and how important it is, you know, achieving maximum value to be customized for those markets. I guess my first question is, do you agree with your peer statements and how are you addressing that? My second question is kind of a big picture, and I apologize if this is dumb.
I'm trying to think, is this new burgeoning market, you know, asset-light, digital-focused AI market, like hedge fund quants that constantly have to reinvent themselves in a more and more efficient market every 5 or 10 years? Or is it something where you're building relationships with project developers and you're innately in, you know, gaining early market share. You're kind of like an Amazon, and you have systemic long-term opportunities because you have such large upfront market share, and you're just a part of the nuts and bolts or the, you know, electrons that connect the entire system. Which is it?
Okay, those are really great questions. Not dumb at all. I think, you know, to the first part of the question, I think, yes, all markets have unique characteristics, right? Optimizing in any given market is gonna look at different kinds of value streams. Many of those, though, break down into fundamentals, right? The fundamentals of energy trading, day ahead or real-time, the fundamentals of balancing services and different speeds of response and the way you get compensated and so forth. Then there's the financial versus physical trades and so forth that can go on. Those are all sort of constructs that are out there.
We address that by looking at the markets and understanding what that means towards the extensions that we need for Athena to do something that might be new compared to what we've already done. That extensibility that you heard about from the tech team is a key part of how we address that. Plus a number of folks on our team that have innate market knowledge in different jurisdictions and so forth. Now, to the second part of the question, the longevity of the relationships we think is important. I don't dispute the idea that we're gonna have to continue to innovate. Those algorithmic performance and trading behaviors are gonna continue to be something that will be important to us.
As we do that, we'll also be looking at how do we expand our ecosystem, trading partners or users who might just have that sort of trading view as opposed to the asset owners who have the asset ownership position. What we do feel is very important is over time, the proof will be that trading an asset means you have to be very physically aware. That situational awareness of the asset, its capability of performing is something that has to be baked into your total solution. Trading purely on some kind of fundamentals with you know some idea of some algorithm doesn't understand the asset itself, I think is not gonna be the winning strategy.
Which I think it makes us sticky with the asset owner as well as with the trading partners who wanna leverage our software to help in some of that, robotic trading processes that I think are gonna continue to drive algorithms forward. Fortunately, we have a strong data science team that can do that. We help to, you know, keep them motivated to keep up with some of these things as well.
I think another big proof point on that is that, and I mentioned it in my discussions around over half of our contracted volume was from repeat customers. It's a real attribute to, I think, what Larsh and his team have built, so. I love your note, by the way.
Let's ask 1 more from the web. Maybe I'll ask a couple. Do you plan to expand in wind energy projects, wind or other forms of solar, renewable non-solar energy? Talk about hydrogen storage. Is Athena able to handle hydrogen storage? This is probably for you, Larsh.
The answer there is that, I think, yes. I think there's gonna be, you know, different opportunities with different asset classes to look at how we can provide that single pane of glass across this whole clean energy spectrum. I think, you know, as Bob pointed out, you know, the monitoring, you know, is something that started with the initial solar industry and understanding that you had to be able to get that data to monitor systems and operate them effectively. We'll continue that as we think about these new assets. You know, as Bill pointed out, though, we'll also be looking at making those investments when we can see the ROI and when there's gonna be that kind of benefit.
From a platform standpoint, we are, let's say, certainly very aware of those asset classes and making sure that we continue to have them in our radar from a product roadmap standpoint.
I think the one that wasn't mentioned in that question that's super compelling is EV and the fact that the EV side, we believe, is the equivalent from energy usage standpoint of the behind the meter market. It's a massive TAM. You know, we've announced some things in the past. We did 1 last week with InCharge. We've done some things with ENGIE, and I think you'll see more of that activity. Wind is interesting. What I find compelling is if you could get the predictive analytics piece that we have for the markets we're in today, you know, just a couple percentage points in that market are huge dollars. I think it's an interesting market.
We've got Matt Tappan, who leads our Corporate Development here in the room with us, and he's always getting calls on a variety of deals, and we'll certainly update everybody as those come together.
Yeah. I'm curious, within the process of a customer building out a site, when do you kind of start discussions with the customer and sign a contract and then therefore, kind of what's the timeframe that you have very good visibility in the growth outlooks?
I'll start it. Bill, maybe you can jump in. I think from a C&I standpoint, if we divide the 2 out, C&I typically will try to do a master service agreement. Take, you know, Walmart and say, "Okay, we're gonna do X amount of stores in Y region." What's interesting is when new regions open for a variety of reasons, whether it's legislatively or lower cost hardware, some of our first customers into those new markets are the C&I customers that we've delighted in other regions. T hat piece, they know the product. We know what to kind of plug and play at their new locations in a new region. I think the other thing they rely on us on is what can be optimized or monetized in that specific new region. The perspective of front of the meter, it varies.
In some cases, we're very early on. We have, as mentioned earlier, you know, the fact that they're coming back to us so repeatedly. They'll open the book of pipeline business that they have and work with us on those. I'd say those can be much earlier, and mid-stage as well. It just varies by developer and maybe the relationship we have with that customer. Typically, once we close a deal with them, though, we move upstream from the perspective of timing on that individual customer. I don't know, Bill, if there's anything else you'd add to that.
Yeah. I think maybe your question was maybe even more specific as to, like, the actual duration as well, potentially. You know, we're starting to transition away from a BTM/FTM kind of distinction. I think it's really size of project is probably gonna be the more relevant metric going forward. I think the smaller projects, i.e. like 5 MWh and smaller, you're really talking about a pipeline duration, you know, from opportunity to sign contract, like 0-6 months. That's really kind of what you're looking at, is as our teams work with the developer to actually design the system and make sure that the appropriate hardware is chosen. That's really underlying this shift out of, you know, the kind of the disaggregation of margin from hardware to services.
That, that's a service that we're charging for now. The larger projects take longer. I mean, you know, if it's a much more complicated interconnection, the system design that the teams have to go through is a much more complicated one, and so you just see a longer time period.
This question is for you, Bill. I understand you're not giving annual guidance, but with more software-only deals coming into the mix and the potential to monetize services more, again at higher margins, how should we think about free cash flow kind of going forward over the next few years and then uses as you scale?
How do I think about free cash flow? I think you know, we're in favor. You know, I think the you know, that's where the EBITDA percentage comes from. I mean, really, I mean, this is not gonna be a business where we're gonna be buying, you know, building factories, buying equipment, owning assets. I mean, there's other folks that are much more efficient than us to do that. You shouldn't expect to see that. When we think about adjusted EBITDA percentages, those are the free cash flow. I mean, there's really gonna be a very small differentiation between those things for this business.
You know, we don't see ourselves being competitive with, say, like a GSAM or Brookfield or somebody like that who can bring, you know, hundreds of millions, if not billions in dollars in terms of, you know, asset ownership. Those are great partners of ours. We've been serving them for a long time, and we would expect and hope to continue to do that in providing services to those. I don't think from that standpoint, you know, we would get ourselves into a position that we needed, you know, or had assets which would make a significant differentiation between EBITDA and free cash flow.
Yeah, thanks for taking my question. Software and services have been flattish at $5 million for the past few quarters. When do you think we can expect to see uptick in those?
I think that you should see that in the coming quarters. I think that for really 2 reasons. 1, I think there have been, you know, effectively, if you go back and say like software and services, what drives those? It's AUM. AUM is increasing, not as quickly as we would like, and for all the reasons that I've talked about, interconnection, et cetera. We're starting to see some projects come through that, so we'll see more services and software revenue coming into the business, which is what's driving, you know, when you think about, like, how do we get to cash flow positive in 2023? I mean, that's not because we're selling more hardware, I can tell you that much. It's because we're deploying services and expectation around that.
Anyone else? Okay, great. Well. Oh, 1 last.
Gotta hold this side of the room up.
This will be our last question. We'll call it after this.
Andrew with Morgan Stanley. Just curious on UFLPA, what percentage of your volumes next year are dependent on panels getting through customs? If, you know, if it's a high percentage, how easy is it to reallocate to maybe standalone projects?
I think for next year, I mean, clearly the, you know, say, the solar monitoring part of the business is gonna be impacted by UFLPA. Has been so far. We expect that to continue. What that means in part is that the projects are slower. It's not that they go away. I mean, that's a really important point, is that we've never had cancellations of major projects. There are, you know, size differences or size changes and that sort of thing, but cancellations are few and far between. I think that's what you would see. The battery part of the equation seems to be less impacted. We've done a lot of work on which is obviously the bigger revenue part to the business. We've done a lot of work on the supply chain, working with our suppliers.
I think part of the differentiation is that in general, the battery supply chain is a bit more sophisticated than maybe the panel supply chain. You know, there's more differentiation between, you know, say, a Tier 1 and a Tier 2 or 3 on the panel side than there is on the battery side. I don't know, Bob, if you have any thoughts on that one.
you know, that's an interesting question, and appreciate that. My perspective is, and we've lived through 2, you know, the Section 201, the Suniva. We've lived through a number of different kind of disruptions. you know, my experience is that it's not demand disruption that we're talking about here. You know, it's deferral. you know, people scrambling to figure out how to refill the pieces of the puzzle that the chain's kind of lost. you know, it's hard to tell right now, but quite honestly, you know, my expectation is that, you know, the growth will occur. It would make it deferred in time. It's hard to tell exactly. but this is consistent with what we've seen in the past. It's not de-bookings, right?
It's simply deferrals.
One thing I would add is, we primarily see it in the utility scale side of the business in the Alliant Energy side. I'd note that the BTM side is approximately 80% or more of what they do. You know, one of the things that's been great from an execution perspective is they've vectored to drive greater growth there, as well as pushing through price increases that fully washed out any impacts from UFLPA delays on the utility scale side of the business. You know, that's great agility and we expect, I think as we've talked to folks on the logistics side, most people are expecting the H1 of next year is when it gets resolved.
Really just kind of the forms and the clearances need to settle out with CBP and the ports to start flowing. We do have just a mix shift that helps us as well.
Okay. Well, I wanna thank you all. Hopefully, you've enjoyed the morning as much as we have hosting you. As Ted mentioned, there's lunch behind these doors. The management team, executive team will be here to answer any further questions. Again, thank you all for your time today.