Hello, and welcome to today's call for GFT Technologies 2025 preliminary results. My name is Philip, and I will be your host today. On the company side, we have presenting Mr. Santos, who's the Global CEO, and Mr. Ruetz, who's CFO and Deputy CEO at GFT. Following the presentation, there will be a Q&A session. To ask a question, you have to raise your virtual hand. Afterwards, I will give you the permission to speak. With that, I hand over the word to Head of Investor Relations, Andreas Herzog. Please go ahead.
Thank you, Philip. Ladies and gentlemen, welcome to today's conference call and our preliminary results for 2025 financial year. My name is Andreas Herzog, and I'm Head of Investor Relations and CSR Compliance at GFT. Joining me on the call today is our Global CEO, Marco Santos, and our CFO and Deputy CEO, Dr. Jochen Ruetz. Marco will begin with an overview of our key developments and highlights of the year, sharing also some light on our successful path in AI, followed by Jochen, who will walk you through the 2025 financials in more detail. Marco will then conclude with our outlook for 2026 and some key strategic takeaways. The full presentation materials are available as usual on our website. With that, I will hand over to you, Marco.
Thank you, Andreas. Good afternoon, everyone, and thank you for joining us today. Let me start with a clear and simple statement. In 2025, we delivered and exceeded our current guidance. This performance is particularly significant when we consider the broader market environment. The technology industry is going through an accelerated and continuous process of paradigm shifts. Artificial intelligence is accelerating software development, changing productivity dynamics, and reshaping how clients approach technology modernization. Given my academic background in computer science, I could not be more excited to seize the resulting opportunities as Global CEO of GFT Technologies. We anticipated the fundamental change early on and crafted our AI-centric five-year strategy to be a driver of this transformation. Therefore, we are not changing direction. We are accelerating execution. Next slide, please.
2025 was defined by disciplined execution and tangible progress following the adjustment of our outlook earlier in the year. In an environment marked by currency volatility, accelerated technology change, and structural shifts in our industry, our priority was clear: to stabilize performance, to strengthen the quality of our earnings, and close the year with improved momentum. We are proud to announce that we exceeded our guidance for 2025 with improved profitability in the second half. Let's take a closer look at our 2025 results. Revenue reached EUR 888 million, representing 2% growth in euros and 5% growth in constant currencies. Our EBT adjusted reached EUR 67 million, corresponding to a 7.6% margin, which demonstrated positive improvement compared to the first nine-month margin of 7%.
The EBT also improved in Q4, reaching EUR 46 million with a 52% margin, reflecting both better operating performance and cost management compared to the first nine months margin of 4.9%. Excluding our portfolio optimization case of U.K. and Software Solutions, our corporations delivered 12% of growth in constant currencies with a 9% EBT adjusted margin and a 7.5% EBT margin. Our main growth markets in 2025 showed strong momentum, with Brazil growing 28%, Colombia 19%, the U.S.A. 17%, and APAC 17% year-over-year. Colombia delivered particularly strong performance, reflecting the successful integration and accelerated growth of our Software Solutions acquisition. In the United States, we achieved solid 17% of growth, reinforcing our positioning in the most competitive and strategic technology environments worldwide.
In terms of sectors, insurance and industry recorded strong double-digit growth. Overall, the second half of the year demonstrated that our strategic decisions and operational focus are translating into improved margins and stronger quality of earnings. This performance reinforces that our transformation towards an AI-centric operating model is not only strategically sound, it is financially sustainable. Based on this performance and our solid financial position, we will propose a dividend of EUR 0.50 per share, unchanged from 2025. Let me reinforce what I said in the beginning. We are leading the IT industry's transformation. We do so because we didn't simply echo the hype of the market, but because we identified the underlying potential of generative AI and its profound relevance for our industry as early as 2023.
To be precise, in June 2023, we kicked off the strategic initiative to bring generative AI into the core business of GFT, our engineering services. We knew that we must have an IP to orchestrate AI and launched the development of the GFT AI Impact product. Only six months after the launch of ChatGPT. Well before AI became the dominant topic across the enterprise markets and the IT industry. Throughout 2024, we moved from spec- experimentation to deployment, strongly integrating GenAI into our engineering services, software development lifecycle, and delivering first clients and project implementations. While other people were building chatbots, we did the hard work, validated use cases in real and mission-critical environments, and built up internal know-how and capabilities. As a result, we published our AI-centric five-year strategy to the markets in March 2025.
We doubled down on the profound positive impact of artificial intelligence, and we positioned GFT to become a true responsible AI-centric technology powerhouse. From there, we accelerated execution. We rebranded Wynxx based on an architectural evolution of GFT AI Impact towards an agentic AI platform. We expanded globally across clients, countries, and users. In our Capital Markets Day of October 2025, we launched our AI modernization offering, powered by Wynxx Legacy Transformer, and continued expanding the platform into new domains. We built AI native capabilities early, formalized the AI-centric strategy, and we are now scaling it. The GFT AI-centric five-year strategy is more relevant than ever. It is perfectly aligned with all the major market shifts we have seen over the last 12 to 18 months.
When we defined our strategy, we positioned GFT to become the best responsible AI-centric digital transformation company in the world. Our mission is to bring the best responsible AI-centric digital solutions, software development, and technology service to every company in the world. Not as a marketing message, but as an operating model. We embedded AI into our DNA, our delivery platforms, in the core process of our software development lifecycle, in the implementation of our partner platforms, into modernization methodology, and into the way we scale globally. Again, I could not be more excited to say that our strategy is more relevant and validated today than ever. The major AI market shifts we are witnessing are not challenging our direction, they are reinforcing it. We developed a comprehensive transformation roadmap built around clear strategic goals, defined objectives, and measurable KPIs.
Our execution is structured across four core dimensions: brand, client, service, and talent, supported by strong group-wide governance, organization, leadership, and performance management. All strategic initiatives are fully in motion. Today, I will touch on some of the key ones: our AI-centric transformation, our rebranding positioning, the expansion of global accounts, Tier one and Tier two clients, and high value-added services and offerings. What I will show you today is how this strategy is being executed constantly and with measurable impact. This is like... Yeah, thank you. Let me now turn to our execution of the five-year strategy, our strategic initiatives, and the tangible impact we are creating with key clients and through our brand positioning. Firstly, driven by our proven AI capabilities, we continue to expand and deepen relationships with large strategic accounts.
During the fourth quarter, we expanded two additional group Tier one clients, each now exceeding EUR 25 million annual revenue. This reflects both client trust and our ability to scale long-term partnerships. In Colombia, our largest clients accelerated strongly with 35% revenue growth and has now become a top group Tier two client. This demonstrated our ability of successfully integrate company acquisitions and to unlock growth through offering cross-selling. In the United States, we achieved a 10% revenue growth with a key major retail banking client, which moved into our top Tier two client category. This is particularly important in a highly competitive mature market where expansion is driven by differentiation and delivery quality. Beyond clients growth, we also strengthened our global positioning. In 2025, we completed our global rebranding across 20 countries, built around a unified responsible AI-centric positioning.
Finally, I am proud to highlight that GFT has been named the number one, the global leader in the IDC MarketScape for worldwide cloud-native core banking implementation services. This recognition, one of the most important in our history, confirms our engineering depth, our domain expertise in financial service, and our ability to deliver complex transformation programs at scale. Let me now turn to the execution of our AI-centric transformation strategic initiative and the tangible progress and impact we are creating with AI. First, we significantly scaled our Wynxx AI agentic platform focused on software development lifecycle and legacy modernization in 2025. It's active in eight countries, serving 92 clients, representing more than 250% of growth in clients year-over-year.
With a total influenced contract value of EUR 70 million, representing more than 700% revenue growth year-over-year. This reflects not only AI adoption, but real commercial impact. A key enabler of this strong acceleration has been Wynxx' multimodal architecture, which allows our platform to leverage over OpenAI GPT, Gemini, and especially Anthropic Claude Sonnet and Anthropic Claude Opus models, which are delivering the highest efficiency and code quality results across our legacy modernization and software development use case. These models are now at the core of our most advanced enterprise AI deployments. A strong proof of this impact is Bradesco Seguros, the largest insurance company in Latin America.
There, the adoption of Wynxx has expanded in 22% quarter-over-quarter, increasing our AI native team from 180 to 220 engineers while delivering a 40% productivity improvement across software development legacy modernization. This clearly demonstrates that AI is not reducing demand, it is driving revenue growth and value creation. Beyond software development, we are also developing an agentic AI platform for credit risk management at a top Tier one bank in Europe, and a large scale agentic platform for orchestration of manufacturing, engineering, and field services for a major automotive client in the United States. These initiatives show how GFT is leveraging agentic AI expertise into mission-critical business process at our clients. We also strengthened our ecosystem.
We are leading the deployment of Anthropic Claude Code and GitHub Copilot across large teams of 50 and 130 engineers, respectively, at two Tier one clients in the U.S. These are also concrete examples of the implementation of AI native engineering teams whose objective is not hand-coding software anymore. Rather, they develop software orchestrating AI code assistance, agents, and large language models. We also launched GFT's global AI native software development center of excellence to scale multimodal and agentic engineering across our organization and clients engagements. It's designed to be tool and model agnostic, allowing us to integrate and orchestrate leading solutions, focusing currently on the AI coding tools Anthropic Claude Code and GitHub Copilot, as well as OpenAI Codex and Gemini Code Assist based on client requirements.
This structured approach ensures focus on the most capable tools on the market, as well as governance, best practice, and scalable adoption while maintaining flexibility and independence in an evolving AI ecosystem. At the same time, we launched our AI modernization offering powered by Wynxx's Legacy Transformer. This formalize our approach to accelerating legacy modernization through AI and establish a clear commercial framework for capturing this massive growing market opportunity. I will come back to this topic later. Finally, we expanded our Wynxx's agentic AI platform into some fronts, its foundation, business process, data intelligence, studio, and marketplace. Position it as a comprehensive platform rather than a single product solution. I will give you more details in the following slides. Taken together, these milestones show consistent execution, scaling the platform, proving impacted clients, expanding into mission-critical domains, building ecosystem capabilities, and productizing AI modernization.
This is how we are operationalizing our AI-centric strategy, not as a concept, but as measurable progress across markets and clients. As I mentioned, I significantly scaled our Wynxx agentic AI platform in 2025. This reflects real commercial impact, not experimentation. Wynxx covers the full software development lifecycle, from story creation and estimation to testing, code fixing, architecture support, and legacy transformation. It's deeply embedded into delivery. Importantly, the platform leverage leading foundation models from Anthropic Claude, OpenAI, and Google. Combining the strongest capabilities of each in a true multi-model architecture. This ensures performance, flexibility, cost-efficient consumption of tokens, and independence in a rapidly evolving AI ecosystem. Scaling the software development lifecycle in legacy modernization was only the first step. Let me now show you how we are expanding beyond software engineering into the broader enterprise landscape.
Through our foundation layer built on this MCP service, governance, orchestration, FinOps, and integration across AWS, Google Cloud, and Azure, we can deploy agentic capabilities into complex enterprise environments. At the same time, our architecture remains multi-model and cloud agnostic, including sovereign clouds and open source model capabilities when and where they are required. The next layer of evolution and expansion of Wynxx is into business processes. We are moving beyond software development into functional, operational, and business domains such as HR operations, resource staffing, project portfolio management, financial risk assessment, compliance, financial transaction monitoring, manufacturing maintenance, among others in pipeline. Earlier, I shared examples of our agentic credit risk solution and our large-scale automotive orchestration program in the United States. They are all becoming new elements on Wynxx business processes. These are not isolated use cases.
We have developed three agents for our own global HR function at GFT, automating workflows and decision support for talent acquisition, which we are bringing into Wynxx to offer to clients as a scalable AI value proposition. In addition, within our SmarAct platform, we are developing agentic AI anti-money laundering capabilities, bringing AI-driven orchestration into financial transaction monitoring and compliance environments. This is a natural extension of our deep regulated sector expertise and reflects a fundamental shift. AI is no longer limited to software code productivity. It's an orchestrator and decision maker of business workflows. This positions Wynxx not just as a software engineering tool, but as an enterprise-enabling platform. We did not stop there. The third layer of expansion focus on data and intelligence.
We are combining data intelligence products and assets and extending Wynxx into industry-specific data domains, from banking, insurance to manufacturing, robotics, oil and gas, and retail. This is where domain expertise becomes critical. agentic systems must operate within regulated frameworks, industry data models, and mission-critical architectures. Our approach combines AI orchestration with deep domain knowledge and sector-specific data intelligence, enabling intelligent advisory systems, all built on a scalable foundation. What we are seeing is a strategic progression from software development to business process to data intelligence. Now the final step. The next layer of evolution is the democratization of Wynxx. At its core, Wynxx integrates software development lifecycle capabilities with business process orchestration and data intelligence, from software development, operational workflows to industry-specific data intelligence.
This platform sits on a strong foundation of governance, orchestration, and multi-model flexibility, supporting Anthropic, all the models, including Sonnet and Opus, as well as other leading foundation models in a sovereign and enterprise-grade architecture. Now, Wynxx is no longer a single AI tool with the launch of the Wynxx Agentic Studio and Marketplace. We are opening the platform for structured extensibility and ecosystem participation. The Wynxx Agentic Studio allows all GFT teams, more than 12,000 engineers, employees, and our clients to build, customize, and integrate agentic capabilities into their own environments in a governance way. The Wynxx Marketplace enables reusable agents, assets, and accelerators to be available, shared, and consumed at scale across all GFT. We are not building isolated AI features. We are building an AI-native operating layer.
This progression reflects our strategic ambition to move from AI adoption to AI orchestration, from isolated use cases to enterprise-scale AI-centric transformation. Let me now connect our AI-centric platform evolution to a concrete commercial opportunity, AI legacy modernization. First, independent industry researchers forecast that the modernization market is expected to grow from approximately $25 billion in 2025 to around $57 billion by 2030, representing around 18% CAGR, growing multiple times faster than traditional IT services. At the same time, 62% of U.S. organizations still rely heavily on legacy systems, 75% of banks globally run on legacy core platforms. These systems struggle with regulatory change, product time to markets, and increasing total cost of ownership. The most striking number, an estimated 250 billion lines of code remain in active use globally.
AI modernization is not a short-term IT initiative. It's a multi-decade structural cycle. The magnitudes of accumulated technical debts is unprecedented. Forbes Global 2000 companies are estimated to hold between $1.5 trillion-$ 2 trillion in technical debts. This is not a technological constraint. Structural bottleneck that hampers innovation, agility, and growth across enterprise. AI can compress legacy modernization timelines from years to quarters. That's what Anthropic recently announced to the market. This is fully in line with what we presented during our Capital Markets Day in October last year. AI is making large-scale modernization more feasible, more affordable, and less risky, unlocking productivity and cost advantage that were previously out of reach.
This combination, massive technical debt, structural demands, and AI-enabled acceleration, defines a significant brand-new growth opportunity for GFT today and in years to come, something that simply did not exist about two years ago. We have already delivered more than eight AI-native modernization project case studies using our Wynxx Legacy Transformer, achieving productivity improvements between 42% and 70%, and up to 95% accuracy in business rule handling, among many other efficiency KPIs. These are not pilots, POCs, and experiments. They are large production-grade modernization programs delivering measurable outcomes. We begin with the modernization strategy itself using generative AI combined with more than 35 years of GFT's domain expertise. We define a structural transformation roadmap aligned with business objectives, regulatory constraints, and long-term architectural goals. From there, we design the cloud-native target model. It's not simply about moving systems to the cloud.
It's about re-architecting them using the best market practices, resilient, scalable, and ready for continuous innovation. Strategy and target design alone are not enough. Execution must be embedded into the full lifecycle. That's why we integrate AI across development, security, and testing to continuous integration and deployment, ensuring modernization delivers not just change, but performance and operating rational control. What makes large-scale modernization particularly complex is that business logic often exists in two places. First, in documented process and operating modules. Second, and more critically, hidden deep within decades of legacy lines of code. That's why our approach combines two perspectives: a top-down AI approach, where we preserve and modernize business rules within the new architecture, reimagining process while maintaining regulatory and operational integrity, and a bottom-up AI approach, where they extract hidden logic, undocumented dependencies, and deeply layered functionality directly from legacy systems.
Many of these rules exist nowhere else but in the code itself. That is where our Wynxx legacy transformer becomes crucial. Wynxx reads, understands, and structures legacy codes, enabling us to restructure complete applications in a modern architecture while preserving and evolving core business rules. This combination, strategic redesign, cloud-native architecture, lifecycle-embedded AI, and agentic code transformation, allows us to move from manual migration to industrialized AI modernization. That's the key difference. We are not just migrating systems. We are transforming business platforms in a governed, secured, responsible, scalable, and repeatable way. To conclude this part of the presentation, let's take some steps back and gain a strategic perspective. All the examples and tangible results presented here underscore our leadership in driving our industry's transformation.
This is not only because we apply Wynxx to become an AI-centric, but because we generally believe and understand the major shifts of AI. We started our AI-centric journey way back in early 2023, ahead of the competition, and we are leveraging on more than 35 years of domain knowledge engineering experience, a clear asset in these transformative times and a claim that not many can take. Jochen, over to you.
Thank you, Marco. After this really important deep dive on our technological changes in the GFT business landscape, let's move to the 2025 financials and directly jump to slide 17, where we see the overall KPIs at a glance. Revenue, Marco already pointed it out, grew by 2 percentage points. In constant currencies, that's 5%. Now allow me to be very precise, 5.3% is the exact number in constant currencies. The order backlog is down by 2%. In constant currencies, up by 1 percentage point. This is a bit weaker than a year ago. When we moved into 2025, maybe you remember, a year ago, we talked about a couple of long-term contracts in the U.S. which were supporting the order book. They are still there, but they are one year younger now.
Therefore, we're still okay with the order backlog as we see it today. Looking at the profitability, EBIT adjusted is down 14%, mostly due to the strategic real-ignment in U.K. and Software Solutions that we talked about in the quarters before. I mean, the bullet points on the right, you see that U.K. and Software Solutions burdened the EBIT adjusted with EUR 14.8 million, while all other units improved the EBIT adjusted by EUR 4.2 million. This even includes the EUR -1 million FX effect. When we look at the EBT, the Earnings Before Taxes, we, of course, see the development impacted by the already mentioned operational challenges. On top, on the EBT level, we had a one-off in 2024 in line with the fiscal court proceedings in Brazil.
A one-off of EUR 9.8 million positive, which could of course not be repeated in 2025. We saw higher capacity adjustments than in the year before. They stood at EUR -13 million in 2025. It was EUR 10 million a year ago. The other way around, M&A effects were smaller. We had less cost from M&A effects. EUR -8.6 million in 2025. That was nearly EUR 14 million the year before. There were some minor virtual share effects. We had a stable tax rate of roughly 29%. Moving forward, slide number 18, let's look at the growth by sector and changes in our client portfolio. When I start on the left, here we see that in the sector industry and others, we saw growth of 14% in the year 2025.
This client group now representing 12% of GFT's overall revenue. Insurance clients grew by 15% of the revenue in 2025. Now, this category represents 17% of the GFT revenue. Last but not least, the big block of our banking business decreased by 2 points to 71% on the overall revenue. Main driver here was the U.K. business, which was shrinking in 2025 and exclusively with banking clients. On the right side of the slide, we see our client portfolio. We categorize the clients in bigger than EUR 25 million, EUR 10 million-EUR 25 million. Marco used the terms Tier one and Tier two, which is the same meaning. Our clients above EUR 25 million or Tier one clients, grew by two new clients. It's now representing 29% of the overall GFT business.
We were able to expand two Latin American clients into this group. At the same time, the clients between EUR 10 million-EUR 25 million , are internally called Tier two clients, went down slightly to 28% because we lost two big ones to the Tier ones. Between the smaller clients above 2.5 million, below 2.5 million, it's quite stable compared to previous years. Moving forward, slide number 19. Let's look at the fourth quarter in detail. On the left side, we see the revenue. First comparing it year-over-year, Q4 2025 versus Q4 2024, we had a solid revenue growth of 3% or 7% in constant FX. Mainly driven by our business in Brazil, Spain and Colombia.
We saw a pickup versus the last quarter, Q4 of last year versus Q3 of last year, of 9%, mainly due to positive business development, again in Brazil, Spain, Poland and Italy. Megawork contributed roughly EUR 4 million in the fourth quarter. On the right side, the EBIT adjusted, again comparing to last year's fourth quarter shows a 9% reduction, which is mainly due to the U.K. realignment that we have been working on. Margin reached 9.4% in the fourth quarter. Comparing to quarter-over-quarter, Q3 versus Q4 2025, we saw a significant improvement of 40%, primarily driven by high margins, good business profitability in Brazil, Spain and the U.S. Margins stood at 7.2% in Q3 and 9.4% in Q4. Moving forward, slide number 20.
Here we look at our performance per business segment. On the left side, the revenue evolution. Let's start with continental Europe, where we are down 3%, mostly due to macro headwinds, particularly in Germany, Italy and Spain. Of course, Software Solutions is in these numbers too. -3% this year 2025 compares to +9% in the year 2024. We had a good growth in Europe in 2024, but the -3% was a bit disappointing from our European market in the year 2025. Looking at Americas, U.K. and APAC, we see revenue grew by 6%, which was 12% in constant currencies as this is obviously all non-Euro countries. By that offsetting the reduction in the U.K.
We showed 6% growth despite the decline on the revenue side in the U.K.. On the right, the EBIT adjusted by business segment. I usually mention at this moment in time, the end of the year, that this is probably not the best KPI for GFT as our two business segments are very interwoven with internal delivery, nearshore, offshore. Still, let's look at the numbers. Europe is stable on EBIT adjusted, more or less the same number, no change to previous year. Americas, .U.K. and APAC is 12% down, primarily to the business decline in the U.K.. The U.K. was EUR -12 million versus last year. That was heavily impacting this part of our business. Moving forward, slide number 21, showing the revenue growth rates of our different markets. As usual, I start at the bottom.
APAC and others, our smallest region, grew by 17% in 2025, but it only stands for 3% of the GFT business. 17% growth is a good result, mainly coming from Singapore, the Emirates, Dubai, Vietnam and Thailand. U.K., often talked about this year already, -29% versus previous year. Of course, most of this coming from U.K.. You might ask yourself, "What's the others in U.K.?" Yes, we separate Gibraltar from U.K., and that's the only additional effect that we have. U.K. and our clients in Gibraltar are both down. Our transformation process is strongly progressing. Our new and very experienced managing director in the U.K. started February 1st and is already with clients today. North America growth 2025 is 8%, fueled by our U.S. evolution, 17% there.
Canada growth was 3%, leading to an average 8% growth for North America. Latin America growth, 22%, strongest contribution coming from Brazil and Colombia. In Colombia, we have a month of M&A insight. We acquired Sophos in February 2024. Organic growth in Colombia was 10%. Overall, good growth on the Latin American side. Last but not least, Europe, minus 4%, driven by a more or less flat Spanish evolution, but a decline in Germany, Italy, and let me mention again, although we don't put it on the slide, in Software Solutions. Bringing me to slide number 22, our income statement. I think we can do this quite fast. Revenue is, of course, unchanged, 2 percentage points.
Second bullet point is highlighting the one-off we booked in 2024 from the Brazilian fiscal proceedings worth EUR 10.58 million. When looking at personnel costs, this is the challenge to be improved, which will improve the margin sides of personnel costs versus overall revenue. The ratio we're showing in bullet point number four, we have to be back to 83% for what we want to achieve in our guidance. Last but not least, let me mention the tax rate. It is exactly 28.5%, as in the year before. That's what we foresee it for 2026. The tax rate will be between 28.5%-29%. This brings me to slide 23, cash flow statement. We started the year with a net cash of EUR -42 million.
We saw operating cash flow at EUR +43 million, which is significantly lower than in the year 2024 for two main reasons. First, we had a lower profitability in 2025 than in 2024. Second, we already mentioned that a year ago, we had very, very positive, favorable working capital effects at the end of 2024, which immediately means it burdens the year after. A lot of our big clients either pay before new year or after new year. In 2024, that was very positive. In 2025, working capital was okay to slightly unfavorable. That was, in the end, pushing the operating cash flow to EUR 43 million. We had hoped for getting closer to 50 in the end, believe me, the money came in in January.
Cash flow from investing activities was mostly Megawork acquisition, which burned us for EUR 6.8 million. Cash flow from financing activities, we had the acquisition of treasury shares worth EUR 15 million and the shareholder dividend. We mentioned the free cash flow adjusted, which is at roughly EUR 28 million. The main drivers why it is below 2024 is the lower profitability. We already mentioned favorable working capital at the end of 2024 burdening 2025, and the not so favorable working capital at the end of 2025, which will benefit 2026, which we will talk about in the coming quarters. Let's move forward to slide number 24. Very quick on balance sheet, really nothing to write home about. The overall total of the balance sheet is down by 2 percentage points. Equity is mostly stable.
Equity ratio is down a bit to 41.1%. Conscious of time, let's move forward to slide 25. Let's look at the left first. Employee number was up 2% throughout the year, comparing year-end to year-end, 11,772 full-time equivalents. On the external contractor side, we see a increase to 1,445. Please be reminded that the company we acquired in Brazil in September of last year, Megawork, works in the SAP field of Brazil, which is a pure freelancer market. The whole SAP market in Brazil is freelancers only. We added 300 freelancers via Megawork. Obviously, the rest of GFT slightly reduced the usage of freelancers in 2025. Utilization in the middle came in at 92.6%.
Let me mention on top that inside this utilization, we had less billable hours than in the previous quarters, which we immediately saw on the profitability side. To the very right, attrition slightly down to 12.2%. Of course, overall up compared to a year ago. This little decline, I look back at 2023 and 2024 reporting, we always have that in the last quarter. Seems like in Latin America, once spring is coming, summer is coming, attrition is reducing somewhat. That's why this explains the reduction to 12.2 from 12.7 at the end of Q3. It's mostly linked to our Latin American markets. My last two slides, we always have two an additional milestone slide on our achievements, and this is slide number 26.
When we look at how did we do in our free cash flow milestone, we wanted to achieve EUR 35 million of free cash flow. We achieved EUR 28 million. I already mentioned on the cash flow side that working capital was not favorable. That was the missing EUR 7 million at the end of the year. Money came in in January, that will help the 2026 free cash flow. Net debt was spot on, 0.8x EBITDA at the end of the year, utilization was also in line with the milestone we gave, 92%. 92.2% to be precise. Now let's look at 2026 for these milestones, look at free cash flow first. We believe free cash flow reflecting higher profitability and more or less stable working capital should be at EUR 40 million.
Maybe with everything I've said on the slides before, that's even a bit conservative. We could exceed those EUR 40 million in free cash flow in 2026. Net debt is expected to be at 0.2x EBITDA if we do not spend on potential further acquisitions. This is assumed in this number right now. Utilization expected to be on a similar level as in 2025. With that, Marco, over to you.
Thank you, Jochen. Let me now turn to our outlook for 2026. For 2026, we are guiding revenue growth of more than 5%, representing approximately EUR 930 million. An Adjusted EBT margin of 7.6% corresponding to around EUR 71 million, and a strong improvement in the EBT margin to 6%, representing approximately EUR 56 million, all figures in constant currencies. In summary, we expect a strong improvement in EBT margin alongside continued revenue growth driven by a deliberative substitution from traditional labor-based delivery towards an AI-native engineering service combined with assets, IPs, and the Wynxx agentic AI platform. Our commercial focus remains clear on expanding global accounts to Tier one and Tier two clients as financial service institutions are increasingly moving towards AI-powered legacy modernization, creating structural long-term demand.
We see this translating into sustainable opportunity across our core markets. We will continue the disciplined execution of our AI-centric five-year strategy and its global strategic initiatives with Wynxx serving as a key differentiator in our offering delivery, client engagements and modernization programs. This includes investments in the modernization of SmarAct into an AI cloud-based anti-money laundering solution to capture new opportunities which will be generated by upcoming regulatory requirements in the DACH region, Europe and globally. In parallel, we will continue investments in AI native assets and IP built on the Wynxx' agentic AI platform, expand our AI native centers of excellence, continue workforce AI reskilling, and strengthen AI infrastructure and platform capabilities across our global delivery model. 2026 is a year of accelerating execution, advancing our transformational to an AI-centric business model while expanding margin, investing in our strategy and sustaining revenue growth.
Let me briefly reaffirm our midterm ambitions for 2029. Our targets remain unchanged. We continue to aim for revenue of approximately EUR 1.5 billion, an adjusted EBT margin of around 9.5%, and a service mix based on at least 50% in high value-added services and AI-native delivery. We plan to further expand our SmartShore delivery model towards a 40% ratio, including AI. These ambitions are fully aligned with the strategic direction we have outlined today. Continued revenue growth will be driven by key Tier one, Tier two clients and global accounts. Cross-selling of our high value-added services and offerings supported by organic growth through our AI modernization and AI-native engineering service, combined with IP assets bundled in our Wynxx agentic AI platform.
Our M&A ambitions are focused on selective acquisitions on high value-added service companies in existing GFT markets. Profitability improvements will benefit from a continued shift towards high value-added services and AI native engineering service, which bring higher margins by design. Further expansion of SmartShore, incorporating AI-centric model and a disciplined focus on scale within our existing geographies, especially our key Tier one growth markets. At the same time, profitability engineering measures and the normalization of SmarAct and GFT U.K. performance are expected to contribute to continuous operational margin improvement across 2027-2029. In short, our midterm trajectory remains intact, the progress we made in 2025 reinforce our confidence in achieving these ambitions. Ladies and gentlemen, to conclude, let me emphasize a few key message. First, we have demonstrated our strategy delivers.
We slightly exceeded our current 2025 guidance. We strengthened the quality of our earnings and improved profitability in the second half, demonstrating operational discipline and focus on execution. Second, we continue to make tangible progress on our AI-centric five-year strategy, expanding two additional Tier one clients beyond the EUR 25 million annual revenue mark. This is a strong proof point of our offering and service differentiation and our ability to scale long-term client partnerships. Based on our successful scale up of Wynxx in the global market, it's moving beyond software development lifecycle towards a broader agentic AI platform, expanding into modernization, business process, data intelligence, agentic studio marketplace, becoming a responsible and secure enterprise AI orchestration platform.
With our AI modernization offering powered by the Wynxx legacy transformer, we are translating this strategic direction into a structured commercial framework to capture a multi-year modernization opportunity at scale. This is a sign that artificial intelligence is expanding opportunities, not reducing them. AI is lowering technical barriers to software development and legacy transformation, making modernization faster, more feasible and less risky. However, large-scale transformation has never been simple and requires more than technology. It requires deep engineering strength and domain expertise. It requires an understanding of regulatory environments, legacy architectures, industry-specific processes and mission-critical systems, and the ability to stay in control, to govern and to orchestrate change across complex systems. This is where we see our role.
We enter 2026 with clarity, confidence and positive momentum, grounded in our AI-centric strategy, differentiated platform and IP, and a strong engineering foundation, and our ambition to lead the AI native era. Thank you very much, and let's go beyond together. Now, Jochen and I will be happy to answer your questions.
Yeah. Thank you very much, guys, for the presentation. As said, we are now heading into the Q&A. If you want to ask a question, I'm asking you to raise your digital hand, which you see on the bottom of your screen. We see the first question comes from Wolfgang Specht. Wolfgang, you can unmute yourself now.
Yes. Hello, good afternoon. Thanks for the presentation. Three questions to start with from my end. First, influenced contract value. Can you give us some more information why you placed influenced in front of that? It's only a part of these contracts, let's say AI, Wynxx related. On restructuring, can you give us some insight where you are currently standing regarding timing for both U.K. and for Software Solutions and what we should expect in further restructuring steps during 2026? Finally, on your order backlog, which is slightly down year-on-year while you're guiding revenues up mid-single digits, does this mean you're looking at a very rich pipeline and expect more short-term deals to come on the table?
That's it from my end.
Okay. Thank you very much for the question. What does it mean, the influenced total revenue? It is, it's a quite simple KPI for us. It is the total value of the project or the service that we sold, that we brought, that we embedded Wynxx as an AI agentic platform. That's the whole services, the whole value of the activity.
It's never just one thing that wins a contract, right? It's usually many factors. Restructuring costs, U.K. Software Solutions, both done, finished. We believe we have the right size since the end of year 2025 in the U.K. and in Software Solutions. The EUR 13 million we spent last year on restructuring, we believe will be far lower in 2026, roughly half of that, EUR 6 million-EUR 6.5 million. Order backlog, yes, we do believe the pipeline is rich enough, exactly as you said, to cover for the revenues that we're foreseeing. The 1% downturn, honestly, this is more statistical delta than it really is a driving force. Order backlog is okay for this guidance.
Okay. Very clear. Thanks a lot.
More questions. Let's jump in.
Yes.
Hello, everyone. Can you hear me?
Yes.
Awesome. Now, well done in Q4, and thanks for taking my questions. First on Wynxx. Firstly, how easy is it to implement Wynxx, and what type of projects do not utilize Wynxx? Is there any category that's clearly not able to utilize Wynxx, for example, or type of customer? I have a question on the market. Clearly, your business ex- U.K. is doing good. What's driving the normalization in your view or also in the discussions that you have with your customers? Is it a normalization, or is the pickup really driven by AI technology, for example, due to the lower project costs or pressure to stay competitive or something similar? Lastly, can you specify the sales outlook for GFT U.K. in 2026? I'm asking to get a better understanding of the different dynamics within the group.
Thank you.
Thank you for the question. Regarding the implementation of Wynxx and how easy it is, it's a super interesting question. As we started that journey way back, as I mentioned, 2023 started GFT Impact. It was the seed of Wynxx. In 2024, we were able to deploy it. We faced lots of challenge of implementation of GFT Impact. It was later on Wynxx, into the clients in order to customize it for their environments, okay, for the clients. Last year, we invested strongly on the infrastructure of Wynxx in order to make it straightforward. That was a very good decision, a very strategic decision that we did.
We invested on the foundation of Wynxx that I mentioned here, we made it much more straightforward. The implementation now for clients is really fast deployments. We evolved a lot on that front. That's the first question, right? The second question, which is the U.K. normalization. What is the driver, right, of the U.K. normalization? We are very, very happy with the new MG of that we hired for GFT U.K., so he's already on board. He's already with clients, already with the teams now. We understand that we are now on the second phase. The first phase was the major change of the organization last year, but now we are in the second phase of this transformation.
We are now on the or let's say stabilization and preparing the ground in order to accelerate. I must say that we are very happy with the results of U.K. generally this year. That's also came positively as we were planning, which is very good news for us.
Let me attach the numbers to it because you asked for it. On slide 36, we anyway show our turnaround situations in a dedicated way in the deck. For 2026... While in 2025, we had revenues of roughly EUR 80 million, in the U.K., we believe we will see the turnaround happen somewhere mid-2026. Overall, for 2026, we expect revenues to be at EUR 70 million. That would be another EUR 10 million reduction coming from the U.K., and that is baked into this guidance. Hope we covered your questions.
That's clear. Thank you. One follow-up question, if I may.
Sure.
That's regarding the projects that do not utilize Wynxx.
Oh, perfect.
Is there any type, or should we assume that long-term or mid-term, yeah, all your projects will utilize Wynxx to some degree, at least?
Well, thank you for reminding for that. Yeah, we are pushing Wynxx into as our core differentiation of GFT and, you know, with a very strong and a good adoption, as you saw in the numbers. However, we have clients and we have global clients that they say, "I have my Wynxx. I have already made a global agreement with Microsoft with GitHub Copilot. I developed my X layer here of governance, and I have my Wynxx." That's fine, and that's fine for us.
We wanna be the best AI native engineering partner of that clients and be the best using the GitHub Copilot, we are bringing our center of excellence and knowledge of that in order to differentiate ourselves from the competition and leverage over this new AI adoption. As a second step, we wanna bring Wynxx later on as an add-on because Wynxx is connected, is already connected with Anthropic, Claude, with GitHub Copilot, Gemini. We can bring that as a second step. That's our strategy. If the client at the very end does not get Wynxx, we wanna be the best AI native engineering service for them.
Miro, thank you. Thank you very much for your question, Simon. Next up is Felix Ellmann. Felix, you can unmute yourself now. Felix, please unmute yourself.
Can you hear me? Can you hear me?
Yes. Yes.
Oh, wonderful. Okay, thank you. Nice to hear these results from you. Congratulations to these. I have one question. With regards to the software segment, many clients in some sectors are awaiting the developments of AI, and they are waiting for GFT signing new orders, looking at the AI field. Some of them even stop their new orders because they say, "Okay, let's look how far the AI developments will go." Do you see any difference in the behavior of clients? Do they sign shorter projects? Do they sign later? Do they even sign earlier, or how do they behave within the fast development we see in the sector?
We see today clients moving forward. We saw last year that several clients or some projects or initiatives, they were kind of in the waiting time, right? In order to understand how technology will evolve and move in direction. We sense now that clients are already, let's say, agreed that AI is part of a future, and they are moving forward with implementation of projects, with development of projects, bringing AI, and even with the implementation of ISV, Independent Software Vendors, Software as a Service. It naturally, we see also more discussions towards build instead of buy.
There are several clients that are moving to us and say, "Hmm, let's review this and maybe build leveraging over the AI." Which by the way, is music to our ears, right? It's very good because we are positioning to be that AI native leader to build. But also implement core banking, Thought Machine, Salesforce, implements other technologies, SAP, and, you know, bring the best to the clients.
Well, on the long term, you could even cut some software revenues from the software vendors of the field by building new software.
Look, it's, I know that's, that is the, it's one of the key discussions, right, on the IT market trend this year. What I can tell you is, as the barriers of creating build are reducing, and if you deploy AI and we can create more, so there is a natural discussion about let's build, right? Let's build and let's evaluate if this functionality, A, B, C, D, instead of I pay a recurrent license for a platform, right, vendor, I can build that. So that's clear. However, I think that core that's my perspective, okay? That's my perspective. I think that the core services, they are gonna be heavily based on SAPs or Salesforce.
By the way, if you look at the sites of one of the largest AI in the market, one of the ones that I mentioned today. If you look at their sites, they are hiring people for Salesforce, for example, right? They are really moving towards the core. They are implementing SaaS. The satellite functions from SaaS, those ones are under discussion in terms of to be built. I think at GFT we are very well positioned because I can build with AI, and I can implement that with the, my SAP partners, and not naturally accelerate with AI for the ISV partners that I have, Salesforce, ServiceNow, SAP, among others.
Thank you very much.
Thank you for your questions, Felix. We have one more question, which comes from Lucas Spang. You can unmute yourself now.
Yes. Hi, good afternoon, gentlemen.
Hello.
I would like to start with the cash flow topic. What you already mentioned that could be kind of conservative. If I take a progressive calculation and take the EUR 35 million you initially planned for last year, take the EUR 10 million EBT on top, and then the EUR 7 million which slipped from Q4 to Q1, I would even come up at EUR 52 million. Just from understanding the EUR 40 million better, what is behind this EUR 40 million? Why just from EUR 40 million for 2026?
We were indeed torn between 40 and 45.
Okay.
There's still one challenge. We still have to pay taxes. The EBT has one cost position not mentioned. On the working capital, we simply took a more cautious stance than you have now done in your calculation.
Okay. On the adjusted margin level, if I compare 2025 to your guidance 2026, it's both 7.6%. Regarding or looking forward, you have this 9.5% for 2029. Why is there no progression in terms of margin in this year versus last year?
Yeah. Two main reasons. The Software Solutions part of GFT is still in investment mode as we have described. We will have kind of the same result in Software Solutions as we had in 2025 on adjusted EBIT level. Again, I'm referring to slide 36 to see the details. The U.K. is back to black in terms of positive adjusted EBIT contribution, but still far off the 9%, even our 9.5%, even the 7.6% that we guide for the year 2026. These are the two challenges we have. Last but not least, we have added roughly EUR 5 million of additional cost for the transformation of the whole company to move ourselves into AI-centric. It's now the time to do this. There we have added investments internally.
When we say investments, it's cost. It's people cost, it's training cost, it's business development cost, it's sales people, to drive GFT exactly towards that evolution, to generate higher margins in the years after. This is the goal of the year 2026, build the foundation.
These costs are rather one time or will these costs stay in 2027 and beyond?
They will stay, but relative to revenue reduce.
Yeah. Okay. Thanks. That's from my side.
Thank you very much, Mr. Spang, for your question. So far, we don't have any further questions in the queue. With that, I am handing over back the word to you, Andreas, for some final remarks.
Well, thank you, Philip. Thank you, ladies and gentlemen, for your time and all your questions. It seems that there are no questions left at the moment. If there are any further questions following this call, please do not hesitate to contact the IR team. We remain for your disposal. With that, we conclude today's conference call. Have a good day. Goodbye