Okay, let's go ahead and kick off. Fredrik, can you just do a thumbs up if you hear me well and see me well, so I know that it's working?
Working perfect.
Thank you very much. Good morning, all, and welcome to this first presentation of a company called Teneo.ai. It's actually our first quarterly report, where we are now Teneo.ai, a fresh start that we did during the summer, and very well executed by the team. It's now all done, and, everywhere we are, online presence, LinkedIn, customer email, et cetera, we are Teneo.ai as a company. Our product is, of course, Teneo, so this is a great start for us to be able to get more recognition, better rankings in search engines, et cetera. Very happy with that, very happy with the team execution on that. Also, very happy with this quarter. Another stunning, exceptional quarter delivered by the team at Teneo.ai. It's gonna be Per and Fredrik talking today.
Fredrik, of course, our CFO, and I am the CEO of the ticker Teneo. Teneo, the ticker, is Sweden's only publicly traded AI company. Some of the highlights of, except for that, is that we also had a record quarter, 90% revenue growth year over year. That is astounding what the product and the team is achieving here. We have three new customers so far this year. We're also closing on new partners to cover a larger part of our TAM. We're just about done with that. I'll be releasing press releases on this in the next few weeks, but that is partners to cover a larger part of the TAM in the U.S. We already have partners in Europe doing that, but we needed to do the same in the U.S.
Very happy with that, too, that we're moving on the right cylinders there. But this last one is really, really tremendous. The growth in our traffic, we have now 700 million API calls as an annualized value into the platform, and I think that's a Swedish record. I just can't imagine that there's any software as a service platform out there that has anything close to this type of volume. That means that we're gonna be doing about 100 million phone calls this year in the platform. It is, I mean, it's mind-boggling, and start to think about the amount of people that would have been engaged in those 100 million phone calls across our customer base. It is truly astounding what's been accomplished with this platform and how it scales and how it manages that traffic.
Extremely, extremely happy with that. See some of our logos on the right side that we're very proud of. Of course, Valcon is the one that came in during this quarter. Slightly different tack on those. It's a partner, but they're also gonna be a consolidation partner, meaning they're gonna use one Teneo to deliver to several customers, and I'll come back to why that's important. This one you've seen before, it's our SaaS journey that started in 2021. I'm not gonna bore you with all the different boxes here, but essentially, where we are right now is we've sold to large customers during 2021, 2022. We understand what they're using the platform for, we understand our unique selling points, our unique capabilities, and we build stories around that.
We're taking that to market, primarily in the U.S. and U.K., because that's where we find that our strengths come to the best, to the forefront, and therefore we also hired quite a senior team in the U.S. and also in the U.K., to attack these really large accounts that we're talking about now. The types of... Let's take an example, American Airlines, these type of accounts that have huge contact centers in the U.S. and the U.K., so the road to a billion. Let's talk a bit about how this works and why that's very important. We aim to get to a billion API calls, which sounded a bit outrageous when we put the goal forward.
At the time, we had 24 billion, 24 million API calls on an annualized basis, and it sounded a bit outrageous, and of course, it requires a lot of phone calls to come into the platform. Why is that important? We have two ways we make money, and yes, I see it's USD and euro on the second row there, so different on the two. But it is because we charge customers either in euro or US dollars, depending on where they are. We also do Great British pounds in the UK. The number is the same, the currency changes. We have two sources of revenue. We have five sources of revenue, but 99% of our revenue comes from these two sources. One is the subscription, what you need to start developing your platform.
This is Teneo Studio. Here, you get a subscription for five people to work and build solutions on top of the platform. We now are launching also a new one here, which is gonna be more expensive for the really large customers that need more than five seats. But 10,000 EUR is our standard pricing for a platform with five seats. Here, you can start building your use cases. So typically, the first 60 days, you're probably transitioning an old keypad IVR, like a Nuance. You're now transitioning that into us, and then you start using the platform for calls. That's when you start generating API calls. API calls are just. So a few, 6 to 7 API calls make up a phone call, and that's just our measure.
The way that we measure, the way most cloud companies measure consumption, is API calls, and we charge a small amount per API call, but of course, when it gets large numbers, that amount gets high. The interesting thing with the API calls is that that's where we make big margins. That's where the real big margins come. On the studio, that is more or less an investment for the first year. The payback starts on the studio license only after year two. However, the API calls, they start making money for us from day one, and the customers that are above a million API calls per month, that's where the 96% plus gross margin kicks in, which is helping our gross margin to, of course, be improving quarter after quarter as well.
So that's why a billion is important, and this is what it looks like for our total customer base. This includes our legacy customers, which are, well, I won't go in, dive too much into that, but we have a few customers who are still on legacy pricing models, legacy delivery models. This chart actually goes all the way to September. The September fell away when I cut it into PowerPoint, I see, so the last date there is September. And the number in September is 700 million API calls if you annualize it, so September times twelve. And we started out in 2021 with less than 200 million. Gets even more interesting when you look at the platform that we launched in 2021, which is the only platform we're selling since the beginning of 2021.
It's the SaaS platform, our software-as-a-service platform. That's the one that also makes the most margin for us. And if you look at this, we've actually done, looking 36 months back, 59 times the API call volume. So, in September here, you can see we're above 500 million and slightly above 500 million, and that, of course, keeps growing. You see the plateaus in this, that's when customers go from, let's say that they're using, that they're developing a new country, or they're developing a new language, or they're developing a new channel. Then nothing happens, then they turn it on, and then it peaks again. So that's why we get a few peaks along the way. But, in general, the trend is, of course, very positive, up, like I said, 59 times, up.
This is what we're taking to market. So what we found also in the first few months of this year is that our unique selling points really, really resonate with the really largest customers, and they take a bit longer to sell to. There are a bit more stakeholders in these customers, but this is where our sweet spot is. What we launched is a program we call Accelerate. This is internal. I'm showing this to you as investors. This is not how we show it to our customers, since it might be controversial, but Accelerate means agentless contact center, and of course, Accelerate is just to make it into a nice name internally as well.
And what it is is we're aiming to be the company that within the next three years, and I'll tell you a bit more about this quarter, while it might be even shorter than that, but within the next three years, we wanna be the first agentless contact center in a top 500 enterprise. So agentless contact center here means that we no longer need level 1 support or tier 1 support, or support level 1 agents. These are the people that take the first phone call, and if you have a more complex question, they'll send you off to level 2 or potentially even level 3. These are the people that will answer questions like, "What's my account statement? Where's that bill? Why didn't I pay the bill? Did I pay the bill?" The more simple stuff that most of phone calls are about.
We see a way now to take that out completely. So we see a way using LLMs, ML, and our TLML, which is our own proprietary product in Teneo. We see a way to eliminate that whole layer in contact centers. So if you look at a large contact center, let's say a few of those really top customers we're addressing, we're talking tens of thousands of people in the average customer. So incredibly powerful thing that we see we can achieve here. These things need to be sold to customers that are sophisticated buyers. Sophisticated buyers in that they know that they can't just use OpenAI straight away. They understand that the hallucinations come with LLMs. They understand that everything needs to be integrated.
You need to have ISO, you need to have SOC, you need to have compliance, you need to have encryption, you need to have an enterprise-grade application managing all of this. They also understand the pitfalls of modern AI, which are, it's very quick to do something quick and dirty and get 80% right. It's very difficult to do the last 20%, and they need that platform that's gonna help them mitigate the risks, both compliance risks, security risks, risk of misdeveloping or risk of miscommunicating to customers, while still harnessing the powers of LLMs, ML, and all the new things coming out of AI, today. So this is our vision.
We're gonna be the ones that provide the first agentless contact center in the next three years, but I'll share a bit more of what happened during Q3, which might mean that this is even gonna move even faster than that. Here's how we go to market. So this is our focus since Teneo AI was launched. We have 30 strategic accounts at the top of the pyramid, 200 accounts in the middle, and then our total addressable market, which are where we have 500 agents, roughly, in customers. The target 200, the 200 is 5,000 to 10,000 agents, and the 30 strategic is where we have 10,000 or more agents. So could be...
I think the largest we've seen so far is seventy-eight thousand agents, and that's of course in the strategic realm then. The strategic accounts we go after with a direct focus, so those, that's our account managers, the three new people we hired here during the summer, very, very senior people that go after these people with what we say in the U.S., Rolodexes, but also, of course, very senior knowledge of our market. We also do ABM campaigns in here. That means account-based marketing, meaning that when the people we need to talk to in these customers, which is twenty to fifty people, depending on some different personas, but twenty to fifty people in these accounts, we have them targeted with ads, with personalized letters, with Google Ads, with Google generic links, et cetera.
We also, in the US, we have a tool which helps us to see that if we send a physical letter, for example, to somebody, and they then go to our website, we'll actually be able to see that they've done that. That's not allowed in Europe, but a great tool in the US to see that our campaign's also working, and they are, so we're very happy with the focus on those 30. The account managers, the account directors, then they also nurse another 200 accounts. These are primarily, they do LinkedIn outreaches, email outreaches, mostly automated, but also some personal outreaches in that space. And here we have lots of marketing campaigns happening. We're also doing lead management together with partners, so we're mapping these out with partners.
This is to make sure that if one of those 200 are gonna replace their IVR, like their Nuance IVR now, or they're gonna they need to go conversational, or they need to cut costs, then we will be able to push them up into the strategic layer and start working them with the full power of the organization. The whole website has been rebuilt to support the buyer journey in these lower parts of the pyramid, the blue and the yellow. Of course, that also helps in the top part, but the blue and the yellow, primarily. So if you look at our website now, it's very much tailored to people that find, "I need to do something." They go out and search, they go to Google, we're top on most Google searches, and they come into our website.
That, that is generating now about 30% of the pipeline in that top, the top two parts of the pyramid. Bottom part of the pyramid, that's where we go after, with channels, primarily. That's why we need three new ones in the U.S., three instrumental ones in the U.S. in Q4. Very close to that. I'll announce that soon. And then we have, marketing campaigns running here, and then targeted campaigns, like, for example, the Nuance replacement that's coming out now. Nuance, which is a company that was acquired by, Microsoft, they have a system that picks up the phone with voice or with keypad. The keypad version of that is in 90% of our target accounts.
Nuance has been communicated two weeks ago that it's end of life in the mid-summer next year, which means the companies need to find a replacement, and we're obviously gonna be there for them and help them replace this. We've done a few of those replacements already, and it takes us about 60 days to help a customer do that, help a customer and a partner do that. We have the methods for it. That's how we go to market. We're addressing the full market. We're very highly focused on the top, since those are the ones that can generate the most API calls. Those are the ones that are gonna get us the next billion API calls. What happened here during the summer, which I wanna, which I touched a bit on during the presentation, now I really wanna step into.
One of our large customers, they used the latest functions that we launched during Q2, which is called Copilot. It is where we have access to OpenAI or other LLMs, but most use OpenAI. Use OpenAI to make the process, but also translate the processes. So if you think of a customer journey, it can now translate that customer journey from, "Welcome, why are you calling us today?" to customer being happy on the other side. If you have that in, let's say, English, you can now translate that into other languages. One of our large customers tried that, and they found it good enough to, in 48 hours, deploy 42 languages in 89 countries. So over a weekend, they expanded from English only into all these countries and geographies.
That means that all their channels, the plain old telephony, IPT, so IP telephony, WhatsApp channels, their chat, et cetera, all those channels, both for support and for new sales, were now in other languages and in other countries. Now, we have something which is, I'm gonna get a bit technical, so, but I'd still want to mention this. One of our core competencies is we do deterministic pre- and post-processing. So we decide what the customer said deterministically, not statistically, probabilistically, which is what ML and LLMs do, so machine learning and large language models. We do deterministically. We first use LLMs and MLs, but then we have our own patented language, which we parse with, and that lifts the accuracy, so we know what the customer is talking about, and then we also know what the answer is gonna be to the customer.
So instead of having, "Well, let's stick some glue on the cheese on the pizza," we're gonna tell you exactly what to do in the right language. So once you have that set up in English, you now convert that to other languages. It just works. Incredibly powerful. It shows the performance, so we scale. It shows that we have that deterministic working, since the customer's satisfaction in this solution is better, but also shows that any customer that has a large home base and needs to manage twenty-four/seven in other countries, this is a perfect choice. So with that, I'm gonna go into very quickly exactly the economics of this for a customer, and I'm gonna show you the last thing that happened here in during this quarter as well.
It was a language piece, the country, geographic piece, but then the other one. We have a system which picks up the phone and lets customers build a system that picks up the phone, understands why somebody's calling with 98-99% accuracy, far better than a human would. On top of that, now, this quarter, we saw two customers go from 40%, which was best before, to 60% of cases on the phone automatically resolved, taking us even closer to that agentless contact center vision. Those calls cost about $0.40. A call that goes through and that gets routed to a human costs about $6. If you move more from the right side to the left side, you're obviously making big money if you have millions of calls per month, which these large customers do.
So the two hundred, so the blue part of the pyramid and up, these are really, really big cost-cutting numbers as well. So that's something that we're taking to market selectively. Cost cutting is not something you wanna bring out everywhere, but definitely very, very important to customers. We're talking millions potentially per month that you could be saving in cost here. Most of the time, level one support is outsourced in the U.S. and U.K., another reason to do this in the U.S. and U.K. Seldomly is that employees of the customer, it's employees of business or business process providers, business process service providers, outsourcers. Also, you get happier customers because you resolve something right away. You call in, they...
The system understands you, there's no queue, and you get your, the reason you called, you get that resolved. The real reason behind this is we've done real tests. We use this extensively in our selling. We've done real tests with real data together with a real testing company, Cyara. We have much higher accuracy than anybody else. So if anybody doubts and says, "Well, obviously, there's no doubt that Google is better than everybody else," well, as you can see here, not in this particular case. So if you were to use Dialogflow today, you would have problems understanding the customer and problem delivering the right answer at the end, which is why those solutions do not scale to millions of calls.
So again, just summarizing my part of this presentation before we go to Fredrik's more deep dive into the numbers. It is Sweden's only publicly traded AI company, Teneo. So if you wanna bet on AI, this is the bet. A record quarter, 90% revenue growth year over year, really, really proud of the team. Done great work here. Three new customers so far this year, existing customers growing like crazy. And we're also closing in the partners to cover a larger portion of the, and build more pipe in the U.S. And tremendous growth. I mean, 700 million API calls. Again, I'm gonna start asking around in my community and LinkedIn and so forth, does anybody have this type of SaaS application with this type of traffic today?
I seriously doubt it, but it is a tremendous feat of performance optimization by the product team. So with that, I'm giving it over to Mr. Törgren.
Thank you, Per. So, let's continue a bit on what Per started sharing with all of you. I mean, this quarter was fantastic in terms of sales growth. And as evident on this summary slide, I think, I mean, just let's just focus on a few of them. But, 90% revenue growth on SaaS API calls year over year is a fantastic number, obviously. And also evidenced by the fact that we are, you know, showcasing our scalability in our revenue model, which is very much based on API call growth, and driving growth with our existing customers, and adding more use cases.
So I think there are lots of proof points in this slide, which summarize a bit high level what has happened during the quarter. I think we can go to the next slide, Per. A bit on the same theme. I mean, the third quarter, twenty twenty-four was a quarter of fantastic growth, especially in our SaaS revenues, and also the fact that we also maintain and keep a stable gross margin and also improve our EBITDA. So basically, shrinking the gap in terms of inflows and outflows in the business. And all, I mean, essentially all sales metrics are record numbers in this quarter, so a very strong development in the quarter.
As already mentioned, I think, I mean, the 90% growth year over year in SaaS API call revenues is astonishing, I would say. Also linked to that one is obviously also that we are growing SaaS ARR with 58%, and also that we continue to deliver strong growth on our recurring revenues, so both on the SaaS side, but also then obviously on the non-SaaS side as well. Also, if we look on the graphs to the right, I think worth mentioning here is obviously our 100% recurring revenues. So basically, we are very much a pure software company relying on recurring revenues only. And that's, I mean, we have had, you know, high numbers on this for a number of quarters now, but...
I think this is evidencing, I mean, that, that we are a pure software company, truly. Also, the 63% share coming from SaaS. For those of you that have been with us since we started this, you know, transformation journey into going SaaS, have seen this number increasing, you know, continuously over the quarters. And this is the highest number we have experienced so far. So a clear majority of the revenue is coming from SaaS, and, basically evidencing that we are growing in the areas we want to grow in. And I think that's also very important as an organization, that we also showcase, what we are focused on focusing on internally. NRR, we already mentioned.
I think that is evidencing also that we are driving growth with our existing accounts, which is less costly than acquiring new customers. And also the fact that we have a scalable model and which also translates into increased volumes and adding more use cases. So very comforting to see these numbers. Gross margin stable, 79%, despite some commission costs this quarter. And also that we have experienced, you know, new customers joining us, and that also put a bit of pressure short term on the gross margin. So a very stable development also on the gross margin side. EBITDA improving, thanks to higher sales numbers and this stable gross margin. Cash position, SEK 32 million in the quarter. We've come back to that.
In reality, the number we had in cash in bank is somewhat higher. As you can see, the adjusted number was 37 million SEK, but we'll come back to that on the cash slide we have later in the presentation. Very solid quarter, I would say, summarize things. We can go to the next slide, Per, please. NRR, key KPI that we are focusing and putting a lot of attention to. During 2023, 2024, our strategy has been to grow on existing accounts. Basically, our organization helped them to grow on existing use cases and also helping them to add new use cases.
As Per mentioned, what we have done here in the recent time is also something we will benefit in the long run as well to make it even easier for large customers to scale across new languages and new geographies, basically. For those of you guys that are not familiar with NRR, I mean, it's basically a KPI that measures how well we manage to grow on existing accounts. A number exceeding a hundred means we are growing in this respect. We also usually put some reference to the number that we record in NRR.
And in Redeye's SaaS Universe for Q2 2024, Teneo.ai was ranked number one in this regard, and I think we have a fair chance to be number one here in the Q3 2024 edition as well. Let's see. But a very solid and strong development, I would say. Next slide, Per. ARR growth. We're growing as SaaS ARR with 58% year over year, and total ARR with 30% year over year. So a very strong development here. And, I mean, growing quarter over quarter as well. So a very strong performance overall, I would say. We can go to the next slide, Per. This one, I think, I want to spend some time on.
Coming back a bit, you know, to the key underlying driver for our revenue growth. This is a breakdown of our recurring revenues, both on SaaS and non-SaaS. As you can see to the right-hand side, it's the API call revenues coming from SaaS and non-SaaS. Obviously, what we can see here is evidence of our execution of our strategy. Basically, we are growing a lot on our SaaS business. The purple part of the graph, you can see that is growing, so SEK 9.3 million in the quarter in API call revenues. Also what it shows is the scalability in our revenue model.
Variable API calls is a key driver for the growth there, so 90% year-over-year growth. And also the fact that we are growing on existing accounts. As you can also see there, the impact from subscription and license revenue has been relatively limited year over year in terms of growth. So the key driver has been, you know, building out API call revenues on existing accounts for this period. And also, if we finally then close on this slide, I would say we are also showcasing that close to 70% of our SaaS revenues in Q3 is linked to variable API calls. So really, you know, coming back again on scalability in our revenue model.
So volume pays off, and that's—it's a bit backwards, heavy revenue model, but once that kicks in, it's a very strong model. Next slide, Per, please. SaaS API calls is also a key metric. And it's very much, you know, linked to also what we have showcased on our SaaS API call revenues, obviously, since it's a volume times unit price exercise. But basically, I mean, this KPI is a true indicator of how our customer's application and usage of them are growing. So the more applications, solutions, covered regions, languages, et cetera, the higher API call volumes. So that comes back to what Per showcased in the earlier slides, right, on how we can scale on an existing account once we have them onboarded.
What we can see here is also that we have had a positive trajectory on SaaS volume since basically inception. So we have grown 59 times the last three years. And also, if we look, you know, short term, our growth quarter over quarter was 32% between Q2 and Q3 this year. So a very solid performance in terms of growth there. Next slide, Per. Even more here on SaaS. So just showcasing again, we grow in the areas we want to grow, so in SaaS. We also show that we have significant growth on SaaS API call revenues, 90% growth.
Again, showcasing the scalability in our revenue model, and also as already said, I mean, close to 70% of revenues is linked to variable API call volumes and revenues. And if we then add to this, you know, when we add even more new customers, the growth opportunity obviously will become even greater. And we also see, I mean, on our existing account, as outlined by Per, there is a pattern in how you add use cases. But what we also see in these large organizations that we have as customers is that there is still plenty of room for, you know, further growth in new areas and so on. So, time is to our advantage, I would say.
Next slide, Per. We are constantly improving our gross margin, and that is obviously linked to our higher API call volumes on SaaS. This quarter, we report stable, somewhat growing gross margins in Q3 versus previous quarters. And as we have also showcased, I mean, that when we acquire a new customer, the commission costs is relatively burdensome. So the contribution short term is relatively low versus a more, you know, mature customer delivering high API call volumes, obviously. And also, we have excluded the reporting of including and excluding commissions due to this reason, so that we are also showcasing the gross margin underlying, excluding commission costs, when that can actually have a significant impact. And there has been some impact in this quarter from commission, obviously.
But also then, I mean, on existing customers, the gross margins are strong, and will continue to improve from current levels as we also continue to grow the API call volumes. So very, very solid gross margin development, despite some commission costs in the quarter. Next slide, Per. Down a bit on our OpEx run rate. We had monthly adjusted OpEx of a bit less than SEK 10 million in Q3. And I think we also guided for that we are looking at increasing somewhat on marketing and sales already in Q2. And that has continued to invest in those areas.
For obvious reasons that we are also showcasing this growth, and get more proof point and evidence on our model, and that we deliver a lot of value to customers. From this level then, 160 million in run rate annualized costs, we see a slight increase the next quarter. Not anything, you know, super material, but a slight increase to this level. And also, as we showcase in the earnings release, we have taken some one-off costs in the quarter, and that's basically to do the organization more fit to our go-to-market strategy. And that has resulted in some one-off cost and obviously also some cash impact from that short term as well.
Part of the costs that we free up from this exercise will obviously be used for also then doing further investment in marketing and sales. That's also why we're guiding on a slight increase in cost versus the run rate here. Next slide, Per, please. Cash position. We had an adjusted cash position end of quarter of SEK 37 million, so that is SEK 32 million in actual cash position end of the month. But then we have also added a bit more than 5 million of receivables that were due end of quarter and was supposed to be paid end of quarter. But these are large customers, and sometimes there are some slight delays in payments, nothing unusual.
And we have received them in beginning of October, so that's why we are highlighting the SEK 37 million. And on top of that, we also have some additional receivables that we have not yet collected but will add even 2.5 million to this SEK 37 million. So you can summarize that, you know, adjusted really, the cash position was a strong SEK 40 million end of quarter. So with that, Per, I'm leaving the floor for you.
Yeah. Thank you very much, Fredrik. I would just open up for a Q&A. I'm gonna try to hit the right button here so I can see all the people. If you do wanna ask any questions about this brilliant AI company, which is the only publicly traded, so you have to buy stock today. The first one who wants to ask a question is Marcus. I'm gonna unmute you now, and then you need to unmute yourself as well. Go ahead, Marcus.
Hey, guys, and congratulations on strong growth, yeah, for another quarter. I have a question regarding the NRR. I mean, can you elaborate some on the distribution of the NRR among your current clients? I mean, it's a very, are you seeing-
So we don't break out-
Increased-
... Yeah, so Marcus, we don't break out individual clients, but for competitive reasons, so since we have quite a few very large clients, we don't really want to sort of get somebody thinking that they should target one of our customers. So we don't really break that out. So I don't know, Fredrik, if you can say something general about that, but-
No, I don't know. Maybe Marcus, we can take this also offline a bit, but not really sure exactly what you are after, if
Essentially, I would think Marcus is saying, "Can you tell us which customers are growing?" Would that be-
No, but what I'm actually asking is-
Okay
... I mean, is the growth very concentrated to one single customer, or are you seeing all clients growing, or how is that distribution?
I would say the SaaS. So, so maybe the way to put it is the SaaS customers are growing.
Yeah.
The legacy customers are not really adding new use cases in the same pace, right, so not a lot of growth on the legacy side.
Mm.
Legacy being the ones who have a different payment model, a different delivery model-
Yeah
... still that we haven't been able to convert.
Right.
Yeah.
Thank you.
Overall, I would say we see, you know, growth in if not all, but most of our SaaS customers, so.
All right. Good, thank you. And regarding the one-offs that you've been taking, can you elaborate some on that? And what can we expect in terms of one-offs, going forwards? I mean, the Q4, you-
Like I say, we don't expect them going forward, but I'll let Fredrik elaborate, but.
Yeah, so I mean, on... We have taken some one-off costs, and to a large extent, I think, we also showcased that in our slide here on OpEx, right? A large portion of the one-off cost is related to staff. So we have basically taken out people, to as we also communicate. So we have taken out the staff, and that's roughly 7 million SEK of the costs. And then we have some other parts related to recruitments and yeah, linked professional fees to doing this exercise, I would say. And on top of that, also, we have, as we mentioned in the earnings release, also cost related to Gecko as well.
So those are the key components of the one-off costs.
Right. Thank you.
Okay, I'm gonna open up the floor to Forbes. You need to unmute yourself as well, Forbes.
Great. Great, thank you. And, well done on strong ARR growth. One question on the financial targets. How should we think about those, especially the cash flow target, now that you write in the report that the new strategy will require some more upfront investments with likely delayed cash flow? Just some comments on that to start with, please.
Yeah, Fredrik, do you wanna comment a bit on that, or?
Yeah. So, I mean, we highlighted in this quarter, right? I mean, we had this a billion target, which we are relatively close to reaching. And also, if we are performing well, and we don't really see any reasons not to continue on the growth path that we see in this quarter for the coming quarters as well. So, I think that is one side of that. And then we have also the 200 million ARR goal target that we have for 2025. And I would say that is still achievable, both of those targets, and also from a cash flow perspective.
I think what we have communicated there is that we're aiming to have one month of cash flow positive from operations during 2024. I would not rule that out, but I think also if you look on EBITDA level, which is a good proxy of the cash flow from operations, we can also see that we need a bit more volume. And that's also why we put a bit of emphasis on, you know, showcasing the volumes and that we are very close, right? But we don't really want to give any guidance on more detailed guidance.
I think, from an analyst perspective, I think if you look at our OpEx costs and gross margin levels, you can see a bit where the inflection point from an EBITDA level is at the moment, and there is a bit to go to be cash flow from operations. Yes.
Great. Thank you. And I'll ask one more question on your visibility on customer rollout plans. On the one hand, your existing customers, we've seen really accelerating sequential growth now for the most recent quarters. And then, what your visibility is through those customers that you hope to get with partners, if that's a bit more unclear, and how involved you are in those talks.
Right. So, if we start with the existing customer base, the thing that we're pushing now across the base is this, geography and language, the way to deploy more languages. Even somebody like a Telefónica Germany also has to cover five or six languages. Germany has Turkish as one of the big languages in Germany, for example, and so even, even local companies have that, of course, and not the least, the healthcare system in the U.K. have a lot of different language needs, which today are being done by interpreters. So, that's something we're pushing quite a lot to include that in the rollout plans.
Another interesting thing is that one of the large potentials in the UK has now decided that that contact center that never really got deployed that they are instead continuing with their old contact center which we are integrating to as we speak. So that's also good news to us that we have a potential real bump and rollout plan there. But otherwise, there's still other use cases, other phone numbers, other entry points to in all these customers still, right? So I don't think we penetrated more than 60% in any one of the accounts. So we still put a lot of our resource to that because that is, of course, the easiest way to get gross margin. On the new customer base, the channels are not really...
The pipeline we have from channels right now is probably six, seven cases. Whereas on the direct side, we have at least 15 that are in advanced stages, so we have five sales stages that are in stage three or higher, so we have several direct engagements still. The channel is something that we're really, really kicking off now in the U.S. primarily, which is where we think that's gonna have the most impact. In the U.S., when you go over thousand call center agents and above, you actually have a very large universe. You even get to things like HVAC companies that we found yesterday, so people that sell air condition, even they have call centers in that size because, well, just because of the size of the market, of course, so.
So we are also going forward, we're gonna start disclosing pipeline numbers. As we present the Q4 report, we're gonna disclose pipeline as well, using both the weighted pipeline and also the amount of deals in the different parts of the pyramid that are in the pipeline.
Excellent. Thank you. That's all.
I don't see any other hands raised here. So, oh, there's one. Niklas, let me see. I'll unmute you here, and now you can unmute yourself, and welcome.
Good morning. Can you hear me?
We hear you.
Excellent. First of all, then, well, congratulations, as earlier said, on spearheading and building this challenging market, at least reaching sort of some form of cash flow break even. Sort of alluding to what Forbes was talking about and the cash flow situation, it seems to me at least, that we're heading into another rights offering, perhaps Q1, something like that, the next year. Any comments?
That would be a board question. I'm not really sure that I would say that that's in the trajectory of the cash, but that would be a board decision. And I think the board would take that decision based on the pipeline development rather than anything else in that case.
Okay, um-
Also to add to that, I think also worth mentioning there is also that we have historically, I think, that is evening out as we add more volumes on SaaS, where we get you know-
Mm
... more recurring, I mean, on a monthly and quarterly basis, cash in from those revenues. But we still have a lot of contracts that we get, you know, paid upfront.
Mm
... for when there are renewals, and we still have a lot of them coming in, in the first half of the year. So I think that is also something to take into account when doing, you know, individual forecasting of cash, I would say.
Okay. Just briefly, if you, if you'll allow me, obviously, then realizing you're putting tremendous focus into the U.S. right now, where the competitive landscape is increasingly, well, it's getting increasingly competitive, to say the least, and especially given the amounts that some of these companies that you're competing with, albeit, they might not have the same platform as Teneo, they're raising substantial amounts of money, in the, I guess, at least tens of millions, potentially hundreds of millions of dollars. Could you describe your thoughts or just talk a little bit about what's happening in the U.S.?
I'll take that.
From a competitive point of view.
Yeah, I'll take that.
... what are you seeing?
So, the way to raise money in the U.S. right now is to put lipstick on the pig. So you do a small application layer on top of the LLM. That is a path that's not gonna work.
Right.
We saw one now that's raising enormous amounts of money, and they've done the virtual assistant at Sonos.
Right.
I happen to be in this office I'm sitting now. There's two Sonos stations. I have a Sonos subscription. I VPN into the U.S. and tried it. It is absolute shite, and the reason for that-
Mm
... is that everybody's trying only with LLMs. And
Mm
... I mean, our CTO, and very much me, myself, as well, and also what we saw with Gecko, with the cases we were engaged with, there's no way to get around the hallucinations. I don't know if you saw, I-
Right
... I reposted. Look at my flow. I reposted a research where they took the same forty queries, the same query, forty times a day to GPT, to Claude, and to Perplexity.
Mm.
And they got hallucinations that varied between eight and 25%. And it's not-
Mm
... it's by design, because transformers are probabilistic, they are not deterministic. The only ones that are doing deterministic is us. So even with Google in HelloFresh, we then had to build on top of Google Dialogflow, because the Google Dialogflow was just spewing out 72% accuracy on both sides of the equation, both the input and output.
Mm-hmm. Mm.
We then had
Saw that
... to put a layer on top of it. So, would it be good to have $4 billion in marketing? Yes. But with this-
Mm
... strategy, I'm not sure it's needed, because we go direct to the customers and show them that we have a solution that works today. We can cut-
Mm
... the cost today. You can play around with all sorts of other stuff for HR and but this is something that works today. So we're trying to downplay the whole AI, LLM thing. And then just on a very general note, we're still not there where I saw that some of the fund managers on LinkedIn were saying, "Now is the time to invest above the infrastructure layer," so above NVIDIA. I don't think so. That LLM thing is, with lipstick on is not gonna work.
Yes.
You need to be a much, much deeper application layer. So-
And the lipstick-
That's why we pay sophisticated buyers.
Right, and the lipstick with that, well, I wouldn't, like, would it sort of be related to a company called, for instance, Maven AGI, recently kind of formed two years ago?
Absolutely.
Do you meet them in the U.S.?
No, we don't, we don't meet them. So in the U.S., our competition today is Google-
Mm
... Dialogflow, and
Okay
... to a certain extent, Amazon Lex. But-
Okay
... they're now more, what do you say in English? But they're now competitors and friends at the same time, so it's-
Ah
... we're actually also partnering up with them. So, we don't see any of those... So for the large-scale contact centers, we don't see any of the new entrants at all. The only one we see that says they can do something similar is Sprinklr.
Okay.
But Sprinklr is a company that says they can do absolutely everything, and then they send it to their thousands of Indians to build it. And we've seen that not work for large companies before.
Okay
... so I'm not sure it's gonna work here either. No, we don't see the new entrants in this.
Okay.
What the new entrants are used for mostly is HR.
Sure
... internal, bots or a chatbot.
Okay.
That's where we see them.
It's a colorful journey. Thanks for your persistent work. It's certainly interesting to follow.
Thank you for your support.
Of course.
Thank you for your support.
All right. Have a great day. Thank you.
Thanks. Any other hands raised? I don't see any other hands raised. So with that, then, Teneo, the only AI company that's publicly traded in Sweden, is shutting down this 90% revenue growth Q3 report. So thank you all very much for attending, and have a fantastic day.