Good morning. I'm Herman Sjøberg, CEO of Ayfie . With me today, I have Hege Nikolaisen, our new Chief Operating Officer. Welcome to this presentation of our financial report for the first half of 2025. Let's jump right to the core of Ayfie's business model. Imagine walking into work tomorrow and you have no access to your company data. No knowledge of your contracts, your product notes, no service manuals, no knowledge of its customers, no email, no intranet, no nothing. How useful would you be? That's what a general AI is like inside a business. Brilliant, but a stranger. ChatGPT is trained on public information. Now, imagine if it can also see your SharePoint, your Teams data, your ERP, CRM, ticketing system data. Today, I'm presenting a very different Ayfie. Over the last 10 months, we have cleaned up, we have refocused, and we have refueled our business.
Ayfie repositioned, that means three things: product and platform, privately, using AI on your data, secure and compliant, go-to-market, a partner-first engine with pricing and incentives that scales, and finally, team and capital, leadership, investors, and operating systems to execute on that strategy. You'll notice that our revenue is lower than last year. In the first half of 2025, NOK 6.6 million versus NOK 8.7 million in the first half of 2024. That's an effect of a customer mix shift while we rebuilt the go-to-market strategy. Instead of chasing one-off sales, we invested to sell and grow through partners. That meant real work and big internal changes. We have replaced our ERP software, we have a subscription management live, we have automated signups, we have automated payments, we have partner contracts, kickback models, and sales incentives. When our partners push us, we can now actually handle that volume.
On the cost and capital side, the elements have changed a lot. The total is slightly down versus last year. Other operating expenses are higher due to intentional product investments. We also strengthened the balance sheet with a NOK 16 million private placement in May. We now have a 50% equity ratio. This is execution towards a partner-driven business model. The lag is visible in the revenue, but the capacity to grow is now in place. To start, I want to show the power in our technology known as RAG, Retrieval Augmented Generation. That's the business Ayfie is in. At Ayfie, we retrieve the right facts from your internal systems and let the AI models work with them, with citations, respecting permissions, and without moving, and perhaps even more importantly, exposing sensitive data. That's something our customers care deeply about.
Let's compare models with, on the right-hand side, and left without Ayfie RAG connected to the large language models. Without RAG, models like ChatGPT or Claude, or whatever you prefer, are limited to public knowledge. If you ask a question like, what is the capital of Norway, it will answer that question based on publicly available data. In that example, with or without RAG, the answer will be Oslo, obviously. With RAG connected to your internal system, it can answer so much more than those publicly available questions. If I ask a question like, how many vacation days do I have left this year? If you have a large language model without Ayfie, to the left-hand side, this will happen. It will say that I don't have access to your HR system, so I can't see your exact vacation balance.
You can log in, you can navigate to the vacation and absence section, select the current year, export, or screenshot it. You will get a recipe, so to say, to execute that task, answer that question. If you take the same large language model and you connect Ayfie to it, this will happen. It will look up in your HR system the exact values that you ask and give you a super specific, super fast answer, exactly what you need in that business process. You can even take this further. If I ask, plan one week off in October, find a stand-in and update the staffing plan. What happens if you have large language models based on public data? It will give you, again, a plan. How do you do it? Check your company's leave policy, choose your preferred week, confirm that with your manager.
You get a plan, you get a recipe to execute on that task. With Ayfie, agentic features connected to your internal data, it will actually do it for you. It will look up what are good dates to have that week off, what's the coverage, who can step in for you in your role, and it can even update your schedule, your calendar, your Slack messages, et cetera, based on what it just proposed. You can simply click approve all and it will be done. That's so much more efficient. If you are wondering why this moment is different for Ayfie, because we have been in business for quite some time, look at this curve. The Retrieval Augmented Generation market, it's compounding towards the tens of billions over the next decade. It's huge, and our story fits that curve.
We were founded all the way back in 2011 by Ed Jakubowicz, the same guy who founded Citrix, with a DNA for infrastructure that scales. For years, we have been building the boring hard parts, connectors into messy systems, permission accurate access, incremental indexing, and even on-prem options. We have some customers that have been using us for quite some time, verifying that technology. It was truly brilliant technology, but we were too early, way too early, and that has been painful for our investors. Large language models came along three plus years ago and changed the market dramatically, putting fresh air into the relevance of RAG. What I have to admit looked a little bit like legacy technology suddenly became a head start, a big head start on the rest of the market.
Cloud, on-prem, data residency, control, that's precisely what our customers want from RAG, and that's what we have been building for more than a decade. Ayfie, we were early, but now the market is finally here. To the investors and alumni who backed us, thank you. The groundwork you funded is the advantage that we scale today. From here, we will measure success by one thing, and one thing only, how many people do actual work using Ayfie. Finally, to scale a partner-first model, you need operators who have done it in the arena before. Sindre Johansen, our new Chief Commercial Officer, background from the Visma Systems, he has built partner revenue from both sides, as a distribution partner and also on the vendor side. He brings the incentive structures, the enablement, and the commercial agreements that we need to have in place to scale successfully with our partners.
Also proud to introduce Hege Nikolaisen. She has helped take 1X Technologies to scale to global awareness and scale. Quite a unique Norwegian story there, supporting major international funding rounds, building operations, and operations are very important, and global brand. I'm very happy to have you both, and with that, Hege, over to you.
Thank you so much, Herman. I am super excited to be joining Ayfie and seeing the company tackle the next phase of growth and to withdraw the potential that Herman just laid out. We are going to acquire customers, and we're going to acquire them through partners. The last few months, we've closed an exciting amount of new partners, and we've trained a total of 150 sales representatives in those companies. Partners have the capacity to cover large grounds and industries, and also taking our technology internationally. We have now approached a good mix of both public and private sectors and have closed in a bunch of new customers that we're also going to show you. Partners will be the core of our growth, and it's personally what makes me very excited to be a part of this adventure.
Their relation and knowledge to their clients and to the businesses and their core markets will give us and them the opportunity to profit on the technology that we've been through the last decade plus. There are strong implications from these partners that Ayfie has a great market fit for both U.K. and Australia. We will continue to explore the expansion in close collaboration with our partners. These partners that we've already onboarded have quite thorough due diligence processes when they accept new technology in their portfolio. That leaves another quality mark for us that we're obviously very proud of. I want to especially highlight Crayon and Canon as our new partners covering large enterprise client base, and Canon also, no, Crayon sorry, being represented in 68 countries.
They evaluate Ayfie as the best and the only available RAG AI solution on the market that allows businesses to extract the benefits of LLMs connected with your own business data. They also see a great fit towards the public sector due to the privacy compliance, and we will get back to that later. We also have Friend as a partner covering law mid-market and public sector, and Deloitte covering public sector and large enterprises. We have Arribatec covering oil and offshore and high demanding security sectors. The pipelines for new partners are filling up, and we acknowledge that it's also quite some work to onboard them. We have therefore developed this four-step process into becoming a partner.
As a step one, we spend quite some time together with a partner validating their technology and seeing if it's a good go-to-market or product-market fit together with Ayfie and the partner. We now offer a much clearer value proposition together with the partner, and we have an attractive pricing model allowing them to easily understand how they can capitalize on our technology, vice versa. As a step two, we integrate them and present them for our system for finance, support, and reporting. We have a transparent pricing structure in place allowing the same conditions for all parties or partners combined with the kickback model divided on three levels. Together with that, we also have systems for billing, finance, support, and whatever is needed to get these partners up to speed and report on what value they're actually creating together with us.
As a step three, we are training a sales team, 150 sales representatives out with the partners' companies have been trained and are ready to scale together with Ayfie in the markets this upcoming autumn. We have a support system internally that can help them answer questions, establishing knowledge bases, and helping them with whatever they need as they integrate Ayfie into the platforms of the clients. Let's look at our new clients that are onboarded. We see that it's being used by actually both non-governmental organizations, private companies, and also the public sector. We're very proud to say that we ensure safe data handling and faster, more reliable work, and we are growing quickly with partners like Crayon and Arribatec and Advansia. What is especially heartwarming for me is the Save the Children case.
It's a case with direct humanitarian impact where we have someone, he's kind of a super user of Ayfie within Save the Children. He says the following. When he collects donations, compliance is critical. He has to go through a bunch of documents, understand regulations, validate that up until the country that the donor is from, etc., etc. What Samuel does today, he logs into Ayfie, he prompts the system, I have this donor, are they compliant, what are the things that I need to be aware of, and Ayfie presents him with a short answer that is a compiled stack of all the documents and systems and whatever Save the Children has across their IT platform. The impact here is actually that he can quickly verify the donor, and the children will get faster money for the children aid will get faster access to money and help.
We also have Bord Energi, who via our partner Arribatec delivered a fully integrated white-labeled solution. Inside Arribatec's platform, they started with the quality system, quickly seeing the value of having this LLM, RAG LLM, and now looking to put Ayfie into their whole technology stack and scale it across the whole organization. At last, we also want to talk about Akershus County Municipality. It's delivered in partnership with Advansia and Crayon. Crayon has a strong presence in the public sector and covering 50% of Norwegian municipalities. That means no tender process is required as the partner is already approved and integrated supplier, enabling a quicker adoption of Ayfie's secure AI solutions in the public sector. That saves us a lot of time.
Now, we've spent 15 years kind of perfecting RAG and nerding out on search technology, and now we see the ideal, I say that it's the ideal companion for powerful AI platforms. Large language models are experts on everything. With Ayfie, they can now also be an expert on your organization, and you can search fully private. I think this is a powerful quote from Sam Altman, also acknowledging the fact that retrieval and memory aren't buzzwords, they're infrastructure. The more context you have, the better these systems get. He acknowledges that ChatGPT also needs an extra layer of tech support in order to extract and really get that stickiness in the product that they have. These big tech companies are facing a new phase of competition in the market. They need to figure out how do we make sure that people keep wanting to use ChatGPT, Claude, whatever.
Microsoft, OpenAI, and big tech leaders are now looking to RAG as the new puzzle piece to really extract the big potential that is currently unutilized in the enterprise market. Needless to say, we are very excited to be in the forefront of this new trend.
Very good.
Cool.
Thank you, Hege. Let me close now with why this moment matters for Ayfie. In the first half, we made a choice to fundamentally rewire how we are going to grow. We now have 150 sales representatives, up against last time I was on set, it was like five internal ones. That's a huge change. We also have all the systems to support that growth in place. That's the big, big difference now from last time I presented there. Yes, the first half year revenue line reflects, unfortunately, our rebuild year. It also reflects a decision to scale what I'm 100% sure is the right way through partners. The machine is now switched on. In the second half year, we are going to be working very hard to capitalize on those changes, turning a great product, which we really have, into broad adoption.
On behalf of Hege, Sindre, and the whole Ayfie team, thank you for backing this turnaround. Ayfie reposition is now live. We have the right product. We are at the right time. Now also finally the right channel to distribute it in. We are ready to grow at scale with partners and with control. With that, let's take your questions. In English, as we have some English-speaking people in the audience.
We have the first question. Why do you think the changes you have implemented so far in 2025 will help Ayfie succeed?
Yeah.
You can take this one.
We already have the partner program in place, and we see we already collect the fruits of it. We have new clients onboarded, and we are filling up the pipeline for new partners as well. In addition to that, we have an extended management team.
Absolutely.
I think these are the three main pillars that will pull us through the next years.
We're already seeing the results of that, right? I mean, we have, I think it was even the last slide, we have increased now from five in-house to more than 100 salespeople.
Sales representatives, yeah.
They have the relationships already. That's a huge difference.
Second question. Why not just buy ChatGPT Enterprise?
Yes. Yeah, I can take that. We get that question a lot. How to explain this? I mean, when you use an AI agent, there are basically three kind of components. There's the thinking part, kind of the engine that creates the answers, and both ChatGPT Enterprise and Ayfie use ChatGPT for that. They deliver the engine, and they even have a part, even a part of their distribution strategy is being implemented into vendors like Ayfie's software. We are not competing there. You can use ChatGPT using Ayfie, and you can use ChatGPT using ChatGPT Enterprise. There is the RAG part, which is where Ayfie really excels. If you have something that thinks, it will need data to think on and work with. Ayfie is about retrieving that relevant information and feeding it to ChatGPT. In that regard, we are not competing either.
If you have ChatGPT Enterprise, you can connect that to Ayfie, and you will get much more value out of ChatGPT Enterprise. We are about the RAG, and that's the main part. I think the only place where we really compete is kind of the user interface. Why did we develop that? Basically because people want fully integrated solutions to buy directly. ChatGPT Enterprise has a per-user license model, quite expensive for some, while we offer a fixed price per enterprise. The user interface is really the less important one of the RAG and thinking part. Those are the components that will really create value, and we are not competing in that regard.
Just want to add also the privacy layer, the fact that law firms now can use powerful language models like ChatGPT or Claude or whatever they prefer, Copilot, and knowing that it's fully private and not to be shared with these major tech companies or not to be red flagged as sensitive content often do.
That's a very good one. Just to build on that, I mean, if you're using big tech AI, they have quite some monitoring and abuse monitoring in place, which could be a problem in terms of sensitive data. In some cases, your data might be read by some people employed at big tech. Using Ayfie, you can build the whole solution behind your own firewall, so to say. That's also a huge advantage.
Next question. What is a good kickback? How much will a partner get when selling Ayfie to one of their customers?
First of all, our new pricing model is fully transparent. You can see it on our website. The kickback for partners is, as mentioned in the stream, divided in three layers, where the first tier one layer gives a 10% kickback to partners who kind of want Ayfie in the stack but don't have the capacity to give them the full support. That's where Ayfie comes in and kind of fills that gap. You have a 20%- 30% kickback for partners that are, I guess, of more of a strategic value for us, where they take the full funnel and the full partner journey, help them integrate in all their layers. They already know their tech infrastructure, and they kind of just onboard the full client. It really depends on how much the partner, I guess, wants to be handling the clients.
Yeah, I just might want to add, I mean, we have been talking about partner sales for some time, but having the whole Visma experience, new investors and board members and employees, I mean, I mentioned it, to succeed at scale with partners, you really need people who have done that before. All that Hege is describing now with these levels of kickbacks, etc., that's quite some work getting in place. I'm just so, it's a very different meeting new potential partners now than it was a year ago when we can present them with this kind of complete partner enablement package, including what we just described with kickbacks, etc.
Next question. ChatGPT was launched three years ago. Why is your progress so slow?
Yeah, and I guess there are two sides to that. I'm also asking myself that question. Why is progress so slow? I mean, we had ChatGPT three years ago, and when we meet clients, they have learned a lot about AI. If we talk to employees that would benefit from using it, it's really not adopted yet. Even really low-hanging fruits that the companies are aware of are not really made available to those employees working in those processes. The market takes some time before getting there, as you saw from that earlier slide on the market. In terms of Ayfie specific speed, it's, I guess, what we just said. It's really been a journey to have that partner enablement package framework that we have now in place with the systems to support it. That's what will make Ayfie able to scale.
It's also interesting to look at OpenAI's go-to-market model with ChatGPT. It was a B2C2B model, making sure that they got a broad adoption on consumer level. When you, on an individual and personal level, are so used to using this tool, you automatically bring it to work, and then you see how useful it can also be at work, opening that perspective to the business market as well. I think it's many things, that there are many perspectives and answers to this question. I think maturing the market is key, and also seeing that what we thought was RAG legacy technology suddenly got hyper-relevant when GPT launched was also something that we had to kind of adapt to and understand how to embrace fully. Many, many different perspectives on this one.
Absolutely.
Next question. When do you assume to have to raise more capital?
We are working very hard now to not have to do that. I mean, we have an ambition now to increase the revenue, so that's not needed. A part of the dynamic working with partners is that it's harder to predict exactly when those sales are coming in. We see now that the feedback we are getting from partners, salespeople coming in, training on Ayfie products, we strongly believe that this will give quite some substantial results moving forward. We are not guiding on that, but we are working hard to avoid it.
Next question. Can you give a practical example of how Ayfie technology has helped a customer more than a normal ChatGPT license would?
Yeah, we already tapped into some of our customer stories. I think another one that is very interesting is looking at Norwegian Armed Forces, where they can't use technology and LLMs that are without that privacy layer added to it. Now working with Deloitte, a great partner of us, and looking to implement Ayfie with such a high security public sector client and requirements really, really showcases that extra value and that extra power that lies in RAG and also that privacy layer. I think that's a good story for our customer to use as an example.
Absolutely. Just to build on that, the dependency on big tech is, well, the dependency on developing the models lies there. We have open source models now that perform quite well, and it's also possible now to deliver quite good AI intelligence from your own infrastructure. This is something that Ayfie builds a lot on. If you want to buy Ayfie and have no data whatsoever leaving your own infrastructure, you can. That's something a lot of our clients and partners value. Very good. If there are any more questions, feel free to reach out to us after, and we will respond. I think you will find a contact us form on the website. Thank you so much on behalf of the whole Ayfie team for joining us today, and have a good day.