We're gonna move on now to ShotSpotter. It's a company we've dealt with. We actually helped them IPO a number of years ago now, so we've known them for a long, long time. I'll say the same thing I said in the last session. I'm really glad to have you here. While the market is punishing my group with the theory that AI is gonna kill everybody, I think it's exactly opposite. I think it's gonna be a great opportunity for most of my companies to leverage it, to deliver services at lower cost and increase the markets they can address. To me, I think we're gonna come out of what has been a depression into a renaissance.
It's the perfect time to sit down and help people who are struggling, you know, to find winners and losers, and I think you're gonna be one of the winners. Let's start. Why don't you each introduce yourselves so everyone knows who's here, and we'll get started.
Great. Thanks, Rich, and thanks very much for having us. My name is Ralph Clark, and I'm the President and CEO of SoundThinking, Inc. I joined the company back in 2010, and was part of the management team that took the company public with Roth Capital in 2017. Very happy to be here.
Thank you very much. Alan Stewart, CFO. I joined in 2016 also. Took the company public in 2017 with Roth as lead. Been here ever since. Love the company, love what we do.
All right. Turnarounds are funny in my space because they often start because of the recurring revenue nature with record revenues and pretty much record earnings. While your stock is down, the company is still basically executing at the highest levels it's been. Let's talk about, you know, what are the basic products you bring to market?
Great. Thanks for asking that question. We're a public safety solutions company. We're primarily known for a technology called ShotSpotter, which is a leading acoustic gunshot detection technology that was invented by our founder, Dr. Bob Showen, who's a brilliant mathematician and physicist who had been doing work at Stanford Research Institute using math to figure out the epicenter of earthquakes. In addition to being a quite brilliant engineer and mathematician, he's also a really incredibly a beautiful human being that wants to make an impact on the world. He had the notion that he could reimagine this technology to be able to deploy sparsely located acoustic sensors to detect and triangulate instances of gunfire incidents or gunfire.
What makes this such a compelling technology is, most people don't recognize this, 80%-90% of criminal gunfire goes unreported in some of our most vulnerable communities. That's the real reason or why for our company. It kinda starts from Dr. Robert Showen and ShotSpotter, that acoustic gunshot detection technology. Since then, we've expanded to other solutions, including license plate reader technology. We've recently acquired a technology that lets us do, I'll call it left of bang weapons detection technology in a very innovative way. I think we've really tried to lean in on the whole physical AI aspect of what we do before it was cool. Hopefully, we're gonna come back to why that's important with respect to the growth of the company. We also have some other solutions as well.
I won't go through each one, but it all really represents opportunities for us to kinda take data, figuring out from the real world and making data usable for primarily law enforcement professionals and security professionals to help make communities and spaces safer.
Let's talk about sort of the scale the company's reached now, the number of communities you work with. Maybe Alan can talk about the revenue scale and earnings scale achievement?
Sure. Yeah. No, we're in our major product, which is the ShotSpotter, we're in over 170 cities. We're also internationally in South Africa, a couple of cities in Uruguay. We're in Brazil, in Bahamas. We are expanding significantly there internationally. In our other products like the CrimeTracer, which is a very large database, we've got several hundred paying customers, but other customers, basically, about 2,000 that provide us information for the data. There's over 1 billion datasets that we are using behind the firewall, the Criminal Justice Information Services firewall. So that's also very important. It uses a lot of AI as well. We provide some additional things such as case management and some additional things like that. That's about $18 million a year for us.
Our total revenue this year in the mid-range of the guidance is $110 million. We're north of that. We're expecting to get about 17% adjusted EBITDA, which is higher than 12% last year. We're growing in revenue, growing adjusted EBITDA. Every single product that we have continues to grow and add more value. And as you said earlier, and I know we're gonna talk a little bit more about, we've been sort of an AI company before it was cool. You know? Four of our products use AI in the algorithms to improve them, and we're already seeing cost saves as well. It's one of the reasons our adjusted EBITDA is gonna continue to improve.
Why don't we talk about, you know, what AI can do and in the context of sort of what the barriers to entry or moats are around your business that make it very more than unlikely someone could replicate what you do with simply using an AI tool?
I'll start and then Alan, jump in, add and correct as appropriate. I mean, I think many of you have heard the notion or idea that AI is only as useful as the unique proprietary data that AI is using. If you have bad data, you're gonna have bad AI. In our particular case, with ShotSpotter as an example, as Alan said, we're deployed in over 170 cities. We have over 1,100 sq mi of ShotSpotter deployed.
We've been doing this for 20 years, so you can imagine the very unique proprietary data set that we have around boom, pops, and bang sounds that are incredibly valuable for us to continue to evolve the model to do a better job filtering out those noise and just preserving and then publishing those instances of gunfire at scale in a very reliable way. In fact, one of the interesting things about the use of our data, it's been court-accepted. We've survived all manner of Kelly- Frye and Daubert challenges, so it's court-tested, and that's very difficult for a new entrant to implement. We complement the AI with a human review to get kind of further, I guess, insight in terms of the data and really kind of filtering out real things that are gunshots.
We use that ability to continue to evolve the model and reduce the amount of noise that's coming into our Incident Review Center. What's really exciting for us is the use of AI or specifically transformer models in the use of our weapons detection solution, SafePointe. What we're doing with SafePointe, which is really unique, is we are providing a weapons detection technology using passive sensors in a very covert way. As opposed to other weapons detection solutions you might be familiar with that have both a checkpoint and a high false positive rate, and you know you're going through a zone because you're getting zapped with frequency to figure out if there's a gun or not.
We're doing it completely passively, and the way that we're able to do that is we're using transformer models to be able to fuse both the speed and direction of the person walking through our magnetic model with a 3D camera, combined with these magnets that are in these bollards that are essentially eight feet apart. People can go ingress, egress through the zone, and we can detect using transformer models things that are. We detect the metal first, and then the AI can help us discern the type of metal it is from the specific dipole that that metal represents. If you think about a gun going through, it creates a very unique signature that we're able to interpret using transformer AI. That's really exciting.
We have a number of data scientists that are involved in that, in that SafePointe project and really are kind of breaking the barrier with respect to having friction-free weapons detection, which is a huge unlock around specific verticals like a hospital that's concerned about patient dignity and visitor dignity, and certainly casinos that are a little bit more economically inspired that wanna get you through this casino and sitting down at the blackjack table and losing money as quickly as possible. They don't want you standing in line, waiting for that. So that's really quite exciting. I'm also personally excited about the, I'll call it the R&D productivity that we're getting from AI. A number of our engineers are embracing it. They're doing vibe coding. They're doing some other things. They're rapidly developing prototypes.
This is particularly valuable to our work with respect to NYPD, where we're basically supporting a number of their ERP systems. That's about a $14 million a year business for us that we've had a lock on now for 20 years. Just driving that level of productivity and responsiveness and supporting that very demanding client is really quite encouraging. We're seeing it. I personally love the, I'll call it the, professional productivity I get from using AI, along with a number of my colleagues.
For example, if I have time to share one example, we've had our VP of customer success that basically developed a multi-agent solution that with a click of a button can basically go interrogate a lot of internal data with respect to customers in terms of usage, ShotSpotter alerts, coverage areas, length of contract, who the players are from the city council to the mayor. They're also gathering outside data as well in terms of press coverage, political sentiment, community sentiment, and they're all pulling together literally in a 30-page report with a playbook based on it. It says, "Okay, based on all this information we're gathering, these are the five things that you need to ask about.
This is what you need to push on." It's almost like a sales playbook that's customized for a specific customer that really drives effectiveness. I mean, that's just one example. We're creating through these various task force that we have a number of these types of applications to go drive productivity of our kind of sales, marketing, and customer success organization.
In the long run, do you think that AIs will be able to drive faster growth for the company as a, as a key, you know, benefit, or do you think it's more of a cost-controlling, you know, leverage that helps you or both?
All of the above. Yeah, certainly. I mean, yeah, it's going to drive stickiness. We've always been a very sticky company. I think that's a thing that I don't think a lot of people appreciate. We're driving like 99% retention from a gross retention point of view, which suggests that, you know, LTV, lifetime value of, you know, 10, 15, 20 years +. From an efficiency point of view, 'cause we've always embraced technology in terms of productivity, the cap that we've spent to create that dollar's worth of ACV has been $0.50-$0.60. I know compared to kind of traditional SaaS software companies, they're typically spending $1.50 to create a dollar's worth of ACV.
Right?
Mm-hmm.
I think they're seeing attrition, or they're seeing their retention rates be at, like, 80%, which would be considered a good one. You're looking at spending $1.50 for a five-year life. We're spending $0.50, and we're driving 15-20 year life. Do the economics. This is a compelling technology solution and technology business, I should say, and we've been doing this. This isn't new for us. We've been doing this for, you know, six to seven years +, and it's only getting better.
Let's talk about the growth opportunities. You know, you've been, on an annual basis, a very steady grower year after year after year. How fast a growth and how large a markets do you think you're going after? How long can you be a good growth company?
Yeah, it's a great question, and really it depends a little bit product by product. If you think about our ShotSpotter, which is the gunshot detection solution, we're in 170 cities. There's multiples of that, even domestically, that need that, have gun violence issues with communities or underserved communities. We're gonna continue to add every year on that. That's gonna maybe not add 100 new miles every year. Maybe it's 50-75 new miles, but we're also expanding internationally. For the first time, we hired a senior salesperson in Brazil because Brazil has about 100 cities that have gun violence at, like, St. Louis type levels, and those of you who know, that's a significant amount of level of gun violence. There's a lot of expansion that's going on there.
I think one of the things that's most interesting for us, though, is the SafePointe, which is the weapons detection solution, which is the one that we are the only one that's covert, or people don't know, as Ralph said, that they're going through that, okay. If you're a casino, like Ralph said, you want people to get in and out, in and out. You don't want them to go in through a TSA type system. That's relatively low revenue right now, a little over $3 million in last year. That could be as high as $50 million in five years, it's gonna be growing about 100% a year. That's a huge TAM. In terms of the addressable market, it's over $20 billion.
That is something that we're really excited about, spending a lot of our time using the AI in terms of building the algorithms that make that solution more appropriate and I think to set the expectations appropriate in terms of what the customer thinks it can do and can't do. What you're doing right now is if you talk about hospitals and casinos, which are the two main verticals we're in, they're right now doing nothing, okay. When you allow them to do something that helps them add some security, it's better than them not doing anything, okay. That is growing significantly. In terms of AI, one really interesting aspect of that, we have invested a lot in terms of AI data scientists and engineers, specifically in the SafePointe. I'll just focus on that real quick.
When we acquired that company in 2023, we had what's called an ARC, an Analyst Review Center. That's the team that actually when you get an alert, they look at that, then they send it to the security team. We've been able to integrate AI into the algorithms that those 50 people that we originally had, even though the number of customers has doubled, has those 50 gone down to 29. When you talk about revenue growth, we're getting it. When you talk about savings and improving the performance, that's real. I mean, that's something that we're expecting to see and continue. The other products that we provide, they're all growing. Those are probably the two that are growing the fastest at this point. Every one of them is, well, four of the six that we have has in...
You know, AI integrated into the algorithms and making them improve every single year.
On the weapons detection side, maybe talk about some of the tailwinds in terms of regulatory actions that have been taken that could drive faster growth and adoption in that space.
Sure. I'll start, and then Ralph can add or correct. I'll focus on two things, the two verticals we're in. Hospitals, for example, in California, there was an Assembly Bill 2975 that was approved last year that is requiring a hospital of a certain size, and there's about 400 of them in California alone, to have a weapons detection solution installed. Originally, they were expecting that to be in April 2027. We think that's gonna be delayed, but that's the first state to do that, okay. We find that other states generally do things what California starts doing, and then they do it. That's 400 hospitals. For example, if we only got 10% of those hospitals or, you know, 40 of those, that adds about $8 million in revenue for us. If we get 20%, 16.
You can do the numbers. That's just California. If you look at some of the other things, I'll shift to the other vertical in terms of casinos. Casinos, Illinois is now requiring all their casinos to have a weapons detection solution. As soon as Illinois did it, Alabama did it. I don't know whether, you know, ultimately Las Vegas is gonna have to do it, but what I can tell you is once you see a particular vertical or an entity start adding something to provide some protection versus doing nothing, it's requiring the other ones that are near them or looking at them saying, "Well, if they're doing it, maybe we better do it." That's why it's such a huge TAM and a big upside for us in the future.
Great. Your guid ance would have you doing best revenue of your career this year, probably the best earnings of your career this year, and throwing off, you know, very good free cash flow. What's the best thing to do with that? You've been acquisitive in the past at times. You've bought back your own stock. How do you think about, you know, the best use of the cash you're gonna generate to support shareholder value?
Well, I guess that's me again. There's really three things you're gonna do from a cash generation perspective or capital allocation. You either do M&A, but we have plenty on our plate right now. We do not need to buy any more companies because everything we have is growing, and some is growing in a significant fashion. That leaves two more things, really. You add more to the balance sheet, which is great. We generate a lot of cash, though. We don't need to add a ton to the balance sheet. What we have historically done is repurchase our shares. Last year, we repurchased about $3 million worth of shares. The year before was higher than that. With our stock price down as low as it is, it's likely that we continue to do that.
If you were to look at us, if you're an investor looking at the total share count, it is basically flat over two years. As we added over 100 people, as revenue went up, it's still basically about 12.7 million shares because even though we're giving out options and RSUs and growing and adding more people, we're repurchasing shares so the shareholders don't get diluted. That is likely what we would be using because we do generate a lot of free cash flow.
Sure. Mm-hmm.
Hi, guys. SafePointe, can you expand a bit on how you price that?
Sure. Yeah. It's a SaaS recurring revenue model in that we're charging $20,000 per lane. What a lane incorporates are-
Lane based?
Yes, lane based. Lanes are basically two bollards that are maximum eight feet apart and includes a 3D camera with a GPU compute at the camera. It's a camera and two bollards. That creates a lane. $20,000 per year per lane. The unit economics are very similar to ShotSpotter in that our cost to... Our full cost to deploy that lane in terms of the bollards, the truck roll, and the camera, and the cabling and all that stuff, it will generally be less than $20,000 such that we break even on that lane on a cash-on-cash basis probably around month eight, and it's going in that direction.
These things have to be serviced and then refreshed with, by lanes?
Yes, there's no moving parts, so it's pretty interesting. Unlike ShotSpotter, which we've engineered to live outdoors and deal with all manner of outdoor aspects and have to get permissions and the like, the beautiful thing about SafePointe is it's mostly indoors and it doesn't have to deal with the weather elements. The permissions are pretty easy 'cause when you sign a contract, that customer's gonna give you the space to establish a lane. Great question. Mm-hmm.
Can you maybe talk a little bit about the competitive landscape and sort of the key areas that you focus on?
Sure. I'll probably say, I mean, with respect to acoustic gunshot detection, that's a category that we invented and we continue to dominate it. We see some noise around the edges from a couple of folks kinda claiming to do gunshot detection, but we still have yet to see any of those be successful at the municipal level. There are those instances of people doing gunshots in parking lots and the like. That's not our business. We're not interested in that. I'm gonna say that's an area we continue to dominate and have a very strong competitive moat around. That competitive moat is basically reflected in the retention rate that we see. You see that. We're not standing still too, I think because we have a very talented engineering group. I should hit on this point too.
We're not standing still with respect to ShotSpotter, which is wide-area acoustic gunshot detection. We've done some tweaking to the sensor, to the firmware, and also the back-end software to now do perimeter-based detection. Here, on a perimeter-based solution, you're not trying to cover, you know, 30 sq mi or 50 sq mi or 90 sq mi in the case of New York City. You're trying to cover a utility substation that might be under threat of attack from a sniper trying to take down the energy grid. There, in that particular situation, we're lining sensors up, and we're not only relying on muzzle blasts, but we're also relying on the supersonic snap of the bullet passing by the sensor to know that the shot is actually coming toward the intended target.
This is critically important when you think about protecting critical infrastructure like substations, embassies, forward operating bases, and even some corporate environments. I think we talked in one of our earnings call about deploying a ShotSpotter solution, including the sniper-based solution that we just came to market with, in a New York headquarters of a very large financial institutions group, located in Midtown Manhattan, where ShotSpotter isn't currently deployed. It opens up a significant amount of opportunity for us there. Although we don't have the perimeter-based solution in our guidance, so that's the potential upside for us this year. With respect to CrimeTracer, again, this is a category that we invented from the previous company that we acquired.
We have over 1 billion CJIS records and documents there, and it took a long time to develop the trust among, you know, thousands of police agencies to be able to give their data to us to kind of put in this pool that we normalize, index, and now we're putting AI, agentic AI and LLMs against that data to make it more usable to people. That's another kinda one and done scenario where we don't have any real true direct composite competition. It is more competitive clearly with our PlateRanger solution. We're competing against other LPR vendors. The good news there is, as competitive as it is, it's also very large and growing very fast, and we just have to get our fair share of that.
When we integrate PlateRanger with ShotSpotter, it really kinda creates a unique value proposition, so we can leverage off of the ShotSpotter competitive moat to improve the success of PlateRanger. Then SafePointe, there, I think the biggest competition is do nothing, and then the next competition would be Evolv. The way Evolv does, for those of you who might not be familiar, it's an intelligent weapons detection solution or intelligent metal detector, if you will. It requires a checkpoint, and it slows things down, and I think we're doing it in a very different way. I think with hospitals and casinos as an example, we have a good competitive position there.
Great. I think we're at the end of our time. I wanna thank you again for coming out at a, what I think is a really great opportunity and timely for people to see you.
Thank you, Rich. I really appreciate it.
Thanks.
Thank you.