I am really sorry, guys. It seems that at least now I have sound. However, Christina, I cannot share the presentation. Could you share the presentation, please, Christina? Could you run the PowerPoint?
Let me do it. One second. Sorry so much.
Thank you. I'm very sorry, guys, for this. Seems to be a new Teams policy that we haven't understood yet. Maybe we could ask our own Teneo and call Microsoft Support and check how it works after this call. I'm going to go ahead and start right away. I want to start by, yeah, sorry, Christina, you need to share the presentation. I guess you're doing that as we speak, right?
I'm opening it. Yes, sorry.
Yeah, brilliant. There are really two things that stand out in Q4. We met a lot of great KPIs. We had great growth. Everything is ticking the boxes of the right financial KPIs. There are really two things that were really, really important in Q4. The first one is that we proved that the Teneo AI model actually works. When we set this model up in the beginning of 2021, the presumption was that as customers would use the software more, we would then also make more money. Therefore, we did not have to scale the cost as we scaled the volume. We certainly proved this during Q4, where we also had our first cash flow positive month. Yeah, you can stay right there, Christina. We had our first cash flow positive month. Also, what you saw is that the gross margin increases as the volume increases.
It really shows that we don't have to scale the cost to scale the revenue. I think that's what we've been saying. Now we've also proven that it actually works. We're very, very happy with this. This is a key thing. We don't have to scale the cost as we scale the revenue. Key thing that we proved during Q4. The second one is that our long march to get the right go-to-market in the US finally came to fruition during Q4. I'll dive in a bit deeper into that. The US is our biggest market. We have the bulk of our revenue there. Since our software is most useful for large companies with large volumes of phone calls, therefore also the US is our most important market. It was very important for us to find a really good way to go-to-market there.
We cracked that during Q4. Very happy with that too. I'll go into more into that. To start off with, some of you might be new. For those shareholders that have been with us for a while, thank you very much. For those new ones that contributed during our last rights issue, also thank you very much. We're very proud of what's happened with this company over the—oops, Christina, you're sharing the wrong screen now. Thank you. Very proud of what we've been able to do here in the last three years since we launched our new SaaS platform. Teneo AI is our new name. That's since this summer. We did change from Artificial Solutions because our product is Teneo. Our customers know us as Teneo.
It is, of course, a lot easier to, in that big worldwide web, brand us if we have the same name as our product. We only brand one name. Very happy with Teneo AI. That has had great success for us as well in terms of rankings on Google, on GPT, on Gemini, and other search engines as well. What I am showing here is the area that we are in. It is customer operations. This chart is from McKinsey. It was published during the summer. It shows that they think that customer operations is going to be the big impact of generative AI. You see that as a blue dot. That means that 75% of the total impact of generative AI is going to be in those blue dots. It also has a high impact as percentage of the spend that goes into that operation.
That's the 40% on the right. Okay, my sound is not good. Let me see if I simply remove my AirPods. Thank you, Stefan, for pointing that out. Let's see. Is the sound better now?
Yes, it works fine.
Okay, thank you. What you see is the customer operations. That is where we are in. That is going to have a big impact with the latest technologies as well, which, of course, we use and have used for the last three, four years. We are an orchestration layer that uses all sorts of AI technologies to be able to receive a phone call, understand why somebody is calling, and automate that phone call. That is what we do. I will move on to—oh, sorry, I will not. Christina, will you move on to the next slide, please? Thank you. The big thing in 2025, which we are all about, is agents. This used to be called, in our case, conversational bots or voice bots. Nowadays, the new terminology, because of the generative AI influence on our market, it is called agents. Now, everybody wants to build agents.
You see that Motley Fool says that that's the big thing. Real-world autonomous capabilities in agents is a big thing in 2025. You see that Microsoft is moving into agentic, launching 10 pre-built agents. Computerworld, everybody's talking about agents and agentic AI. The interesting thing with that is that we are way ahead of the curve. If we move to the next slide, Christina, you will see that we already have 17,000 agents. We have code snippets that do autonomous things. You would converse, you talk to it. It might go to Salesforce, update the record. It might find an invoice. It might open a support case. It might send you a new license key. That's an agent. That piece of software that does that is an agent. We have 17,000 such agents already live built inside of Teneo.
These agents do 25 agent interactions per second. This is the biggest agentic AI platform right now in the world because everybody's starting to build these. We've been building them for a long, long time. Of course, the big thing that changed is with LLMs. The agents became much more effective. We've also raised the automation levels in our customers thanks to generative AI from 40%, which was best in class a year ago, to 60% in 2024. Our agents do 1 billion interactions per year. That's something that we had as a target. That's the 1 billion API call target that we had, which we also broke through in Q4 of 2024. Christina, with the next slide, please. Here's what we do in a nutshell. This is from one of our customers. This customer has 6 million phone calls per month.
The customer is Microsoft. Out of those 6 million phone calls, 60% of them are automatically resolved. If you look at the first left slide here, the system, Teneo, picks up the phone, understands why you're calling. "Hi, welcome to Microsoft. How can I help you today?" You say, "Well, my Teams is muted for everybody, including the presenters. What should I do?" The Microsoft support agent will go out and find the answer and come back and talk to you in natural language and solve this issue for you. 60% of the time, that's solved automatically. 40% of the time, the agent just collects information and sends that information back into an agent. The agent now gets into their agent assist platform.
They get all that information about what the call is about, who's calling, what are they talking about, what product services from us do they have, etc. It results in 100% happier customers. It is very profitable for us with more phone calls, but it is also very profitable for the customer with more phone calls. As you see, a phone call that's actually picked up by a human agent is an average cost of $6, whereas a phone call picked up by the agent, the AI agent, so not the human agent, is $0.40. If you then do the math, 6 million times 60%, 3.6 million a month, are now $5.60 cheaper. That is a big, big saving for the customer. It is revenue for us because every single phone call now is about 98% gross margin in these customers.
That is why we have been focused a lot on growing our existing customer base. Christina, if we go to the next slide. The reason that this works for us better than it works for anybody else is the accuracy. We have tested this together with Cyara, which is an external testing company, which is the biggest in testing natural language agentic AI platforms. They basically acquired all the others. Now they are the world leader by far. They took it upon themselves to test our system against a set dataset called Banking 77, which you can find on Hugging Face. It is natural conversations between a bank employee and a bank customer. Those have been transcribed by people, and the meaning has then been taken out of those transcriptions by humans. That is what is 100%.
We tested this with Google Dialogflow, which is the closest competitor today in these large accounts. That's usually the incumbent together with a company called Nuance. I'll come back to that a bit. 74% Google Dialogflow, something like Microsoft CLU, which we also use in our system, but we enhance that 70%. Amazon Lex, which you might have in a speaker at home, for example, 0.89, also used in Amazon Connect. We are at 0.95. The difference between 0.9 and all the ones below 0.9 is that 0.9 is where an outsourced human call center would be, 0.9, 0.91, and we're at 0.95. It's even better. It's cheaper, of course, but also better than outsourcing to a call center in, let's say, Malaysia or India, which is quite popular as well. Christina, next slide, please. We had a record quarter in Q4.
I really, really think the big number to me that's important is the 135% NRR. So many software SaaS companies have had a reduction in NRRs. That's net revenue retention. Basically shows that customers like us and grow. Most SaaS companies have not been growing last year in their NRR, and they've been below the 100% mark because the parameters they use are employees or seats or something akin to that, like Salesforce, for example. How many salespeople do you have? If you have less salespeople, then obviously you're going to get less revenue. Our model is all about having less people. Obviously, we do not charge for that. We charge instead for these API calls. The fewer people that are in the call center, the more revenue we get, essentially. Very happy with 135%.
That means that our strategy of focusing on these large accounts has worked, and it continues to deliver. We broke what I call the $10 million ARR mark. Now, dollars, of course, it's just about $10 million. It might be $9.9 million today, but SEK 104 million ARR in Q4 2024. Also very proud of that. That was one of our prime targets. And we were cash flow positive. Really, really happy with the Q4 results. Christina, if we go to the next slide. This is, though, the biggest thing that happened in Q4. This is the thing that really changes where we're going. We have been working on top of Genesys in all these large accounts. All the really large implementations have Genesys as a contact center. What's a contact center? It is all these phone lines.
You have lots of phone numbers that are connected to a central system, which is cloud-based nowadays. It is in the cloud. That system can route to Fredrik, to Christina, to Joan, etc. It can reroute, and it can have voicemails, and it can do things like put people in queues. It can also have some intelligence about which queues to use, etc. That is called a contact center platform or CCaaS, contact center as a service, which is the big thing today. The biggest player in that market by far is Genesys. Most of our large implementations are on top of Genesys. Therefore, we have been working quite hard to get Genesys to realize that, hey, we can have a better customer experience if we also work together with Teneo.
We had a big breakthrough in that during Q4 and are now going mostly to market this year. In 2025, we'll focus mostly on Genesys. We're going together with them to their largest accounts. We have an account list from them, which is their large accounts that we are working with them. We're at their kickoff in about a week and a half in the US and presenting us to their salespeople. More importantly is that their biggest reseller, NWN Carousel, we've also signed up with, and we're also at their kickoff in Las Vegas next week. Very, very happy with what we achieved in Q4 together with Genesys. A Genesys implementation is much better together with us. Genesys used to sell a product called Nuance, which is the one that Microsoft had, which Microsoft has taken end of life.
It is a Microsoft product, which is now end of life. You can still buy support until end of 2027, but it is very expensive. Basically, what we do together with on the Genesys accounts is we go there and say, "You need to replace Nuance. You can replace it within the cost envelope you have in your budget. You can now move in with Teneo." We have a package for that that we launched during Q4 as well. This is the global leader. Lots of customers across the world. They are a $6 billion ARR company, and they are also going public during this year. Very, very happy with the development there. Also, we press released.
That was in Q1, though, but we press released recently that we have an employee to run the relationship, and he used to be the general manager of Genesys Cloud, which is the product that we're talking about here. He used to work there, Genesys, and obviously, that helps us tremendously as well. Very happy with the breakthrough in growth to market in the US. This is going to help the salespeople tremendously because we're obviously a small company. Genesys is not a small company. Very happy with that. Yeah, sorry, Christina. You need to click again. I keep clicking on my PC here, but it does not work. If you break this down, there are really two companies that we started working a lot with during Q4. It is AWS Connect, which we had a relationship with since before, and Genesys.
We did a few proposals together with AWS Connect being part of their marketplace that you see here, but also on the Genesys AppFoundry, which you also see here. Those are the two big pictures. In the Genesys part, we have two partners we're working that we're focused on. It's Kenway and NWN Carousel. Both of those also signed up during Q4. In the AWS Connect ecosystem, we also have two partners, AppDirect and Intervision. We're doing slightly less on the AWS Connect in 2025, simply because the opportunity on the left side, the Genesys side here, is very, very big. That is totally going to be our focus for 2025, working together with these large Genesys accounts and helping them to take both cost savings and customer experience to new levels.
Very, very happy with this as we move into 2025. This is what we're focused on. This is where we're moving to, and that's how we're going to really increase the footprint in the US. Also in Europe, but again, primarily the US. Christina, next slide, please. Hit one more key there, please, Christina, so I get the next thing at the top. There we go. How's our revenue model working? We do charge for seats. That's for the development platform. Typically, a customer that has a lot of automation and has millions of phone calls, long term, they will have three people working in the platform. Three FTEs working in the platform. In the beginning, they might have 15 because they need lots of people to work on setting up the conversational flows.
In the beginning, they might need a lot more seats. We charge them for the seats as they're building the solution. That's the 10,000, it's actually EUR 9,900, USD, or GBP that we charge per month for this. As they develop the system, which takes roughly 60-90 days, they then take that live. When they take it live, they start paying per API call. This is when we deploy what's called our endpoints. There we charge $0.009 per API call on average. That's the average revenue that we get for these API calls. If you look at the bottom, you see the analyzed API calls, and you see this is accruing. It adds up in this chart, right?
The API calls that you see here in Q4 2024, that reaches just about a billion on the left side. That's our total. The SaaS API call volume has really grown, overdoubling year over year. It's actually even more than, it's 234% up year over year, which you see in the right graph. Let's take that to the next, and this is our most important metric. Let's go to the next slide, Christina. One thing I promised when we delivered the Q3 report, since we're doing a lot of investment into sales, it's costing the shareholders a lot of money. We want to share how we're doing in that. The sales cycles are quite long. They will be shortened by the breakthroughs that we had in Q4, but they're still quite long sales cycles.
What we're showing here is that the pipeline is growing. The pipeline that we have is only with large, we're aiming only for large customers. Customers that are like the customers we already have. A customer today that we already have might have a total potential of $3 million, $3.5 million, $4 million a year when they're fully built out. When they're at 100% using our system for all voice calls, probably also for chat, they're probably $3.5 million-$4 million a year of revenue to us. What I'm showing here is only the subscription part. Only that $9,000, $9,000, $10,000 that they pay per month for the five seats. That's what we forecast in our system because that's what we measure the new salespeople on. The API calls, we measure what we call our customer engagement team on.
Then we put the other team on. Once we sold it, we put them on to increase the API calls as quickly as possible. Therefore, from a sales perspective, what you see in our pipeline is only those 10,000 times 12, 120,000, essentially. It is actually 9,000, 9,000. So EUR 118,800, it would be a deal that we do. That is how we would mark that in our sales system. Even if the customer has a potential of $4 million a year, we still mark it as EUR 118,800 or USD in our system. If you then take the pipeline stages, let's take the first pipeline stage, which is a qualified. That means we know the customer has a problem. We know they have a budget, etc. And there is a willingness for the customer to do something here. We still have not proven it is our right solution.
It is a 20% weighting. So 118,000 times 20%, that is EUR 24,000. That is how that would show in our pipeline. Customer we are engaged with, we know they have a budget, they have a problem. We still have not proven it is our problem. That would be worth EUR 24,000. If we then look at our 12-month pipeline, this again, only the subscription part, EUR 461,000. That has grown 72% since October 2024. It is EUR 461,000 of weighted pipeline for just the subscription part, not the API call. Very, very happy with this. I mean, this means that we are touching the really large customers. We have seen the ones that have a problem. A lot of them have the problem of needing to replace Nuance. Very, very happy with where we are moving in this space. Sorry, Christina, can you hit the next?
Some operational highlights before I move over to Fredrik. We met two of our financial targets that we set in the beginning of end of 2021, beginning of 2022. We needed to reach an annualized volume of 1 billion API calls. That means one month times 12 is a billion. We reached that during 2024. We also showed that we can be built positive from a cash flow from operations during 2024. Again, adding new customers, adding more revenue is just going to increase that profitability. We are very happy that the model works that way, that we forecasted, or rather that we hoped maybe when we set the model into place in 2021. We had a 300% almost year-over-year growth in the SaaS API call revenues. That is the revenue that has that really high margin and 142% SaaS ARR growth in total.
That's the total revenue. We also took one new logo in Q4, totaling four then over the last few, over 2024. That's a company that's going to use our product as an OEM. That means that they basically build a service which they'll sell to many customers. That's a way to also offer this opportunity to save all this money and make a better customer experience to smaller customers than just the very large enterprises. We signed renewal agreements with BPM, with Widiba, and TIAA. TIAA is a fairly large bank that has a very good implementation of our voice capabilities for banking. It's a very interesting use case. BPM and Widiba are now going to be building the same, actually. Those are, of course, two banks in Italy.
Maybe 2026 will be the year when we start marketing bank as one of our big verticals as well. We also then, and thank you very much again to you on the call who participated in this. Of course, you see also that the share price has held steady after this, meaning that this was beneficial for our shareholders. Also for me, of course, who could not participate in this. We brought in a total of SEK 60 million. There are several reasons for this, why this is important, but the most important reason is, of course, the investments we are doing in the US. Since then, we have already signed up for sales and pre-sales people, one pre-sale, three salespeople in the US who are working under Michael Kenney that now runs the US new sales. Michael is ex-Dropbox, ex-Symantec, very successful in billing markets for those companies.
Of course, with Lee Kayne then, who came from Genesys, this is going to be a very powerful team hitting the US during this year. Very happy with the development. Europe, of course, we already have a mature pipeline. There we are in negotiations with customers, which at least in the next few months are going to come to fruition. A bit longer in the US, but in the summer, we are definitely going to see the first deals coming out of this pipeline development. Very happy with Q4 again, very happy with 2024. Most happy with the breakthrough together with Genesys and NWN Carousel, Kenway. With that, I'll hand it over to Fredrik. Fredrik is muted, Christina.
I need to mute him. Yes, one second, please.
Oh, of course, you need to go back into. Very sorry for this, guys.
We will call the Microsoft Teneo Bot and ask why this is afterwards.
Can you try, Fredrik, or not?
Yeah, now Fredrik cannot mute himself. No, he could not.
Okay, now he should be able to. Sorry, yeah?
There we go.
Oh, yeah, perfect. Thank you. Thank you, Christina. We can move to the next slide, Christina. Okay.
One second, please. Sorry. Sorry, sorry, sorry, sorry. One second, please. This is the first time I do this, so I am trying my best. Okay, now.
Thank you, Christina. I think we have already seen the numbers a bit on Per's previous slides. I will not talk too much about that. I will also come back to the numbers in the coming slides. I think, I mean, we can summarize in a very strong quarter in Q4. I mean, year-over-year, obviously, very, very strong growth in all metrics.
If we also look quarter over quarter, a very strong performance as well versus Q3. Per mentioned a couple of metrics that we are very happy to meet. One of them is obviously that we are cash flow positive on a monthly basis in December, which was also one of the financial targets we set up a couple of years ago. Also towards the direction of the company and the strategy that we have been driving to get to that position. Very pleased with that. Obviously, also gross margin, excluding commission, at 87%. That also showcases the operational leverage in our model and the fact that we also see the impact trickling down to bottom line from getting a very solid contribution from the growth that we experience in Q4.
The difference versus Q3 this year is very visible in the numbers, as you can see in the report as well. We can move to the next slide, Christina, please. On this slide, lots of metrics and all covered in the quarter report as well. I will just conclude that, as I already said, I mean, we experienced fantastic growth in our SaaS revenues and significantly improved the EBITDA. All sales metrics are at record numbers. A very strong development in the quarter. The shareholders on the line here and the stakeholders on the line that have been with us for a longer period of time knows that if we look on the graphs to the right-hand side, we can see that we are experiencing 72% SaaS revenues. We started this journey beginning of 2021, where we had zero revenues, more or less, in this space.
We have grown from zero to 72%. Also during the year, we have increased quite significantly on the proportion coming from SaaS and basically because of growing our existing customers. The same goes also with recurring revenues. When we started this journey in 2021, we had a hybrid model with professional services and then old software model, basically. Now we are a pure software company, evidenced by 99% recurring revenues that we keep for the next month, for the next quarter, and try to grow them consistently. Also, just to mention a few numbers here, obviously, SaaS ARR growing 142% year-over-year. Total ARR, also then including non-SaaS business, also growing 65% year-over-year. Very, very solid performance, I would say.
We will come back to NRR, but obviously, as Per already mentioned, this is driven by our focus on existing account and growing them and also showcases the operational leverage from our business and revenue model and also the benefit for our large customers. Already mentioned gross margin, excluding commission, increased from 75% last year in Q4 to 87% this year. Also with the increased volumes on sales, we also keep control of costs. That also trickles down to that we have a very much stronger EBITDA this year, minus $1.7 million versus $5.8 million last year in Q4. Very pleased overall with the numbers. We will also come back on the cash position, which is very strong here also following the directed share issue that we executed in the beginning of Q1. We can move to the next slide, please, Christina.
Coming back just shortly on net revenue retention during 2023-2024, the strategy has been to grow relatively cost-efficient and focusing on existing accounts. We also see that materializing in having customers that benefit and see a value in what we provide. The more they use us, the more they usually also save and also get a much more better customer experiences. For those that are not really familiar with the NRR KPI, you can basically say that it shows how well we manage to grow our existing accounts. A number at 100% plus means that we are growing in this aspect. As Per mentioned, a number of customers in the SaaS space have been showing declining numbers in this respect. To keep and also improving here during 2024 on this metric is very strong, I would say.
I know Redeye, I think Fredrik as analyst that is covering our stock with Redeye, they also have this SaaS universe for SaaS companies. I think in Q3 2024, Teneo ranked as the number one in this metric. I would actually be quite surprised if we are not number one when they also measure Q4 2024. A very strong development on existing accounts, basically. We can move to the next slide, Christina. API call volumes, and then especially on SaaS, since that's where we are driving the business, is also showing a very strong growth. This is basically an indicator on how our customer applications and the usage of them are growing. The more applications, solutions, covered regions, languages, etc., the higher API call volumes, basically. As you can see here, comparing Q4 2023 versus Q4 2024, we grew this three times.
We also experienced 47% quarter-over-quarter growth versus Q3 2024. Very pleased with that. As Per mentioned, this line only showcases the SaaS API call volumes, but combined with our non-SaaS customers, we met our financial targets of $1 billion annualized in API call volumes in one month. Very pleased with that also, that we could meet that target as well. We expect existing accounts to continue to grow. We also have a number of very interesting discussions with some of our high potential accounts for future expansions on our existing accounts as well. Coupled, obviously, with what Per also mentioned on our reinvestments in new sales, especially in the US. In short, very good and solid numbers. We can move to the next slide, please, Christina.
Briefly, I think SaaS ARR, obviously, as we said, 142% year-over-year growth. Also, I mean, similar direction as with API call volumes. We are increasing quite a lot from Q4 and then continuously over the year with a relatively high spike in Q4. Very pleased with showcasing this growth on ARR. Can move to the next slide, Christina. This is actually one of my favorite slides. I know it can be a bit crowded, but I think this summarizes a bit a number of things on our business model and revenue model. This basically showcases the breakdown of our recurring revenues. It also showcases the evidence of our execution of the strategy. I think that is what we work with every day in the company to drive in this direction and push for even more growth ahead.
It essentially showcases that we are significantly growing our SaaS business. It also shows the scalability in our revenue model. As you can see to the right-hand side in the dotted area, it showcases that our variable API calls is the key growth driver. You can see the purple or dark blue, that one has increased significantly if you look year-over-year. That is showcasing a year-over-year growth of 294%, which is a fantastic number, I would say. It also showcased that we are aligning with our strategy, that we are focusing on growth on existing accounts. That is also why you see that the subscription or license revenues have less growth than the variable part. Also coming back then finally on operational leverage in the model, you can see that more than 75% of SaaS revenues in Q4 is linked to variable API calls.
Very much what we are focused on, what we have been focusing on is also materializing. I think from a management perspective, I think this is very satisfying to see and hopefully also in agreement with our shareholders as well. We can move to the next slide, Christina, please. Quickly on gross margin, solid performance also on the gross margin. As you can see, we show stable but growing gross margins in Q4 versus the previous quarters, I would say. As Per mentioned also on pipeline, when we acquire a new customer, the contribution is relatively low because we also have sales commission that will materialize at that stage. That is why we also showcase the gross margin excluding commission because it can have a significant impact in one single quarter.
Also, as you can see, it's not a super significant deviation, and this will also decline as we are growing the sales volumes, obviously, as well. Important that we showcase that we get bang for the buck on our investments in our existing accounts. A very solid gross margin also here in Q4. We can move to the next slide, please, Christina. On this slide, you see the OPEX run rate, so Q3 versus Q4 2024. This is a bit how we measure and follow up on our investments. As already said, we improved EBITDA quite significantly year-over-year. We have costs under control.
I think also for some of you that also have been on some of the investor calls for the directed share issue, we also explained that now we are in a position where we actually can steer the investments we are doing because we are at the inflection point where we are close to break even. We can also take decisions whether to not go for growth. I think now, I mean, the intention now with the strategy that Per also communicated on going after Genesys, etc., that will drive some costs in the first half of this year. We have, as Per mentioned, started to recruit people.
That is basically to invest in further sales and marketing activities since we also see that we have such a unique position now to benefit from actually growing even faster with more focus on new sales. That is why we also want to guide somewhat. I think in the earnings release, we said we're slightly going to increase. To be more specific, this will roughly mean that we plan to increase costs with approximately 10% on a run rate basis to go for new accounts during the first half of 2025. There will be an increase in costs. That is also, I mean, intentional investments since we also see that we have a strong development on our existing accounts and also then to capture the market opportunity we actually see now in the US. Very interesting investments going ahead.
We can then move to the next slide, Christina. Cash position. We had reported cash position end of December 2024 of a bit more than SEK 18 million. Since we are working with quite large customers, they tend to sometimes push payments over the quarter ends. That also happened for us now in Q4. Nothing really strange about that. That is what most large customers or companies do. We have collected SEK 7.3 million of those receivables early in Q1. We have then added the SEK 7.3 million to this cash position. Coupled with that, we also brought in gross proceeds of SEK 60 million from the directed share issue that was communicated here at the beginning of February. Adjusted cash position with these adjustments, we had close to SEK 86 million in the cash and bank.
With that, I think we feel very confident for the coming year and also looking forward to pursue and continue on the strategy that we have, basically. With that, I'd like to hand over to Per again.
Thank you, Fredrik. Christina, just as a summary, same slide I showed at the beginning, the two biggest things, we've proven that the model works, focusing on large and existing customers has given us the possibility to be cash flow positive. That means if we add a few more of these really large customers, that's obviously going to impact the revenue a lot. We don't need to scale the cost with the revenue. That's really what we proved in 2024, and especially in Q4. We also proved that, or we also broke through in the go-to-market in the US.
For those of you who have been with us for a while, we did try initially and quite successfully made a partnership with Microsoft. However, Microsoft was not successful in the contact center space. It was very good for our product team, but not very good for our sales team, the Microsoft partnership. You may remember we were partner of the year in 2022 since we were driving a lot of the AI revenue within Microsoft. However, Genesys is the leader in this market. They are the winner. Aligning with them now in 2025 is really, really important. Amazon Connect is the runner-up, and that might still give us a lot of, especially, I think in 2026, that relationship is going to mean a lot too. Basically with that, I would like to open it up for questions. I am really sorry that we lost 10 minutes in the beginning.
Obviously, we can stay on for those if there's more questions. Christina, when they raise their hand, you're going to need to unmute them first, and then they can unmute themselves. We have a question from Fredrik. Christina, there you go. Now, Fredrik, now you can unmute. There you go.
Thank you, Per and Fredrik. I want to start with SaaS ARR growth, which has accelerated substantially in the two last quarters. Could you perhaps elaborate a bit on the drivers behind to help us understand if it's sustainable?
Right. The SaaS ARR growth is all about when the customer has built the first sort of foundation, they add on languages, they add on new channels. Maybe they start in support, then they add customer service, etc., etc. It's all about how they grow in the existing customers.
What you see in the ARR growth in SaaS is pretty much just existing customers that have been with us for at least two quarters. Those are the ones that are impacting that. We add new customers, then we will get more ARR growth, of course. We have two really large customers where we still have very large potential to grow. Those are the ones we focus the most on this year. We do expect continued ARR growth, yes, definitely, on the SaaS platform. Maybe also to comment on that, the customers on our legacy platforms are mostly customers in Europe, where there are still some issues regarding the European AI Act, where large customers are moving that, trying to move the responsibility for that onto vendors such as us and saying that if we breach the AI Act, you take the penalties.
Of course, the penalty is being a percentage of the revenue of the target. Let's say a large telco in Europe, that's quite difficult for us to swallow. Our focus is SaaS ARR. Our focus is the SaaS, and our focus is U.K., U.S., really, for that growth.
Okay, great. Regarding the pipeline, that as far as I understand, is only the subscription revenue. Is there typically a strong correlation between the subscription revenue and the potential revenue from API calls down the road?
Yes, there is. The salespeople and also the channels, the partners, they make more money on the customer that has the 1 million phone call threshold per month that we use, which is also 1,000 agents, essentially. Those are also our only target accounts. Those are where Genesys are present, right? The 1,000 Genesys doesn't really sell to smaller customers.
They are the enterprise solution. Yes, there is. You would expect the deals in the pipeline, you would expect all of them to be able to generate at least 5 million API calls per month. That takes 6-12 months, right? From deal to fruition, that takes 6-12 months.
Okay, great. Finally, from my side, a lot of new partnerships during the quarter, as you have talked about a lot, of course. What are your expectations on those for this year and for the mid and long run?
Already now, we have a very large pipeline together with these. Basically three or four accounts with each of these partners that we are highly focused on already now. I expect that during the summer, we are going to see real results where partners actually are doing the deals.
They may be through AppFoundry, that's the Genesys marketplace platform. They may be through Microsoft Transact, where we're also present. We're definitely going to see partners do those deals, so sell those implementations. We did have a partner sell already in Q4, which is the first, right? That's BESL, their first customer.
Great. Thank you very
much. That's all from me. Maybe, Christina, I'll just grab from the chat here, Steer chat. Given you reached analyzed 1 billion API calls in 2024, do you have a new target for 2025? We don't on the API call side. We're setting ourselves an internal target as we speak that I don't think we're ready to communicate until the board also has approved that. We're basically looking very at growing the ARR quite a lot this year. That's going to be the focus for this year.
It is not going to be on the API call side. It is going to be on the ARR side, which basically are the same thing. Still, that is the target we are looking at. The competitive landscape, there is more and more. Some of what we call traditional CAI competitors, conversational AI competitors, those would be the Cognigy, the Kore.ai, etc. We have seen that they are losing share. They are losing customers. Nuance is going end of life. That impacts quite a lot because if you call somebody and it says, "Press one for press two and three or four on the phone call," then they are most likely using Nuance. Ninety-five percent of them would be using Nuance in the large customer space. That changed a lot. If you look now, who is the incumbent that we have to prove we are better at than they?
It's either Nuance, which is quite easy, or it's Google Dialogflow, which typically is a bit more time-consuming to prove. With these benchmarks we did with Cyara, we can prove those as well. We don't really see all these LLM-based startups. I call them lipstick on a pig for a very good reason. You put a thin layer of CX or customer experience on top of an LLM. You're still going to have hallucinations. You're still going to have the data security problem. You can have all the issues with LLMs. There's a lot of money going into that, a lot of companies coming there and waving their marketing dollars at us. We don't really think that that's something that we compete with at this point. Primarily Google Dialogflow. The $200 million ARR target for fiscal 2025, is this still achievable?
Yes, we're still on that target as well, Albin. Definitely during 2025, we're going to have a month where we reach that SEK 200 million ARR. Sorry, sorry. Not dollars. Sorry. SEK. But SEK 20 million ARR, that's definitely our aim for this year, yes. Any other questions? I don't really see the hands. Christina, I think you see if there's any hands.
I don't see any others, but it's being the chat if there is someone that is not seen. I don't see anyone's hand raised.
Here's another one. About accuracy, how big is the risk of competitors catching up? It's actually quite amazing. The two things that are quite amazing is we deliver those API calls with a very high gross margin because we don't use a lot of compute, very efficient in what we call the endpoint where we deploy the solution. That's one thing that's fascinating.
The other one is this TLML, which is a patent of ours as well, the Teneo Linguistic Modeling Language. The TLML has been trained on all the conversations we had in the early days of this company, so think 2010 to 2020. The way we broke that up, we do not take the text, we take the understanding of the text and break that into our database. We are the only ones doing it that way. That is what delivers this accuracy. Google has tried, the two biggest implementations they have are Verizon and AT&T. In AT&T, we do one part, they do one part, they failed, we succeeded. In Verizon, they now do 10,000 calls a month. We do not really see that they are catching up on that either because of the accuracy.
We think that accuracy is something we're still putting a lot of focus on, making sure that we develop that all the time. We don't think there's a competitor in the weeds quite yet. Nuance customers, would you say there's any other provider apart from yourself that is competing for them? Yes, definitely. Pretty much everybody is trying to replace Nuance. Many of those are Google Dialogflow derivatives. You build a product on Google Dialogflow, it is often sold as a white label, so it's a component of another solution. We see a lot of companies going after that, but not in the large enterprise space. In the large enterprise space, we pretty much only see Dialogflow, which is what Genesys presented to customers during 2024. We already have one joint proposal with Genesys in the large account.
The sales guys realized that they get a better bang for the buck with us because they get with TLML, they also get a stickiness they don't get with Dialogflow. I think that there are others aiming for it, but I think we have a real edge. By the way, we also are the only ones who have replaced large-scale Nuance already. The first one was Telefónica in Germany, but we've done several of these large implementations since that. Taking Nuance and Genesys and building something great is something we already have done quite a few times. Okay, Albin maybe I might have sent some new comment earlier. Could you clarify what you meant by saying ARR growth won't come from the API call side? No, no, sorry. ARR growth will definitely come from the API call side, definitely.
Maybe what I was trying to say is that there's two customers we have. They have very big potential on the API call side where we are closing in on getting them activated to use it in a much larger sense than before. The rest is going to come from new customers. New customers we sign up that start using the API calls. So shares about growth momentum in Q1 2025. There's been so many restrictions on what I've been allowed to see and say, but I know Fredrik has some views on January. Do you want to comment anything on that, Fredrik?
I think we were not really guiding, but I think we also said that during this call, I mean, we don't really see anything that says that we will not continue to grow.
I think still existing accounts will be the key here in Q1 as well. Linking a bit to Albin's questions as well, API call growth and ARR and linked ARR growth will be the key driver for Q1, coupled with obviously also additions from new customers. Essentially, we are very positive, let's put it like that, for 2025. That also includes the Q1 2025, basically.
Okay. Again, I apologize for this, but a very good quarter, great momentum in partnerships, which is what we've been striving for. Of course, we've proven that the model works. We now just need to add a ton of API calls, Albin, to reach that 200 million. As we do that, the gross margin will also deliver good results. Steer chat , thank you very much. Thanks all again. Sorry for the technology issues.
We'll sort it out by calling our Teneo and asking how it should be sorted out. Thanks all.