The Laigo Story. One Idea. Two Friends. All In.

A note before we begin. On 30 November 2022, OpenAI introduced ChatGPT to the public. For many, that day marks the beginning of today's AI boom. By then, Laigo had already been founded. Its story begins four years earlier.

Winning Holger Over

It is May 2021. Two old friends are sharing a beer on Holger's terrace overlooking Lake Constance. Michael describes an AI system that can read, recognise and file documents on its own. The first setup works. It does not have a proper name yet; his small team calls the prototype “TransFritz” for now. Michael wants to turn it into a company. For that, he needs Holger.

To Michael, Holger is the best systems architect he has met in more than twenty years as a technology entrepreneur. Holger has run his own IT company since 2003. He listens carefully, asks thoughtful questions and examines the idea from every angle. But the enthusiasm his friend has hoped for does not materialise. There is no commitment that evening. Michael knows Holger well enough to understand why. When he gives his word, he has thought it through.

The two have known each other since 1993. For several years, they worked together in Switzerland. Both were competitive athletes, rowing together in pairs, fours and eights, and competing for several seasons in Germany's top rowing league, the 1. Ruder-Bundesliga. Those years taught them the value of their friendship and of each other's word. Holger lends a hand wherever he is needed, whether the problem is a complex system architecture, a broken coffee machine or something at the rowing club. His solutions are practical, carefully considered and technically sound. If Michael would trust anyone with the technical responsibility for a company, it would be him.

On the water, they learned to rely on each other. When strength begins to fade during a race, a promise made back at the landing stage is worth little. What matters is their shared determination, the next stroke and the trust that the other will match it. That trust holds this evening too, even though Holger needs more time.

Michael does not let go of the idea. To understand what drives him, and why he holds on to TransFritz despite Holger's reservations, the story goes back to a hot day in Munich.

An Idea That Stayed

On 1 August 2018, Michael is visiting Apple Germany in Munich with CoSoSys. Despite the heat outside, the old building is pleasantly cool. He takes a photograph, unaware that he will later want to pinpoint this very day. It is the day an idea takes hold and refuses to leave.

After the meeting, he joins Ms Fach, who handles press relations for his company CoSoSys, for lunch in a beer garden. They talk about Endpoint Protector, data loss prevention software designed to protect confidential company information. A new AI feature can already identify whether a text is source code and which programming language it uses. Michael tells her how well it works. As he speaks, the thought takes him further. What would become possible if AI could do more than recognise the form of a text and actually understand its meaning?

The possibilities would be enormous, especially for protecting sensitive data. A confidential contract would stay inside the company, and Coca-Cola's recipe certainly would. General terms and conditions or a presentation for customers, on the other hand, could be shared. But could a system reliably recognise these differences and help people handle documents safely?

Later, Michael discusses the idea with Roman, his co-founder at CoSoSys. He wants to add more AI features to Endpoint Protector, perhaps even develop a dedicated AI engine. But the idea does not become a project. It recedes into the background.

It never disappears completely. When Michael leaves CoSoSys in 2020 following its sale, he takes it with him. For a while, it remains in the background. He knows how much work separates a compelling idea from a product people can rely on. Then his circumstances change, and with them the possibilities. Just when Michael is occupied with other things, the old question returns. This time, it is harder to set aside.

Jerome Walks into the Office

Following his exit from CoSoSys, Michael initially focuses on developing his other companies. Meanwhile, the pandemic defines 2020. He wants to help his newly hired employees through the crisis, a commitment that takes both energy and money.

In December, Jerome walks into his office. He has experience training AI models and is looking for work. Michael hires him at Bauner Consulting. Later, they tell the story of this beginning with a smile. “Jerome walked into my office looking for a job in AI.”

They start with a specific task. Scanned documents should become readable, receive meaningful filenames and be stored in the right folder. A Raspberry Pi does not have enough processing power. Mathias, who moved with Michael from CoSoSys to Bauner Consulting, works with Jerome to build a more powerful AI computer that is still affordable.

By spring 2021, the complete process works. It is an experimental setup, not yet a finished product. But now Michael has something to show his friend on the terrace.

A Meeting by the Lake

In the summer of 2021, Michael's wife meets Yvonne while walking by Lake Constance. Yvonne, a mathematician who graduated with distinction and previously led a small AI team at Bosch in Stuttgart, now lives by the lake and is looking for flexible work. With a young child, the pandemic makes that difficult. Michael's wife introduces her to him.

Talking with Michael about AI and documents, Yvonne recognises a challenge she can put her experience to work on. Later, she will say of that meeting: “That was when he had me.”

A founding team begins to take shape in Michael's mind. Yvonne brings expertise in mathematics and AI, Mathias brings hardware and security experience, and Holger brings systems architecture. Michael himself wants to provide funding and help shape the direction. After many conversations, the four decide to start a company together.

The Idea Gets a Name

There is still about a year to go before the company is founded. In a shared experimental “playground”, the team explores intelligent document processing. There are breakthroughs and setbacks. What product will emerge remains an open question.

In the summer of 2022, Michael thinks about Lego. The same bricks can become a ship, a house or a truck. He wants to apply that modular principle to AI. Existing functions should be combined in new ways to create suitable systems for different customers. All that is missing is a name.

On a walk along Lake Constance, “lago”, the Italian word for lake, comes to mind. It connects the place where the founding team came together with the sound of Lego and the letters “AI”.

Lego. Lago. AI. Together, they become Laigo.

The name brings together a modular building kit, the power of artificial intelligence and the lake where it all began. Laigo is a reminder that individual building blocks can become something of their own, and that sometimes a walk is all it takes for an idea to become a shared journey.

A Familiar Feeling in an Unfamiliar City

In July 2022, Michael visits Skopje for the first time. The city is new to him, yet feels familiar. Through two former Bauner Consulting employees, now freelancers, he already has connections in North Macedonia. Now he wants to find out whether something more can grow from them.

Skopje reminds him of Cluj in 2004. Young people, well-trained IT talent and a desire to build something in their own country. Back then, he founded CoSoSys there with Roman and grew it into an international company. The memory gives him courage, but also puts him under pressure. Was Cluj a model that could be repeated? Or is he seeing in Skopje only the story he wants to see?

He knows reliable people there and understands how to build a technology company across borders. But freelancers must first become a team, contacts must become an operation, and an idea must become a company. Friedrichshafen remains Laigo's home. Skopje will reveal whether the success of Cluj can be repeated.

Keeping the Data Safe

At the end of July 2022, before ChatGPT's public launch, Yvonne, Mathias, Holger and Michael establish Laigo GmbH in Friedrichshafen. Four years after the idea in the beer garden, it becomes a company.

Laigo wants to process sensitive documents such as contracts, payslips and patient records securely online. Holger develops a protected document workflow. Certificates secure transmission and access; keys and files are designed to be deleted after handover. Yvonne takes responsibility for the AI models, drawing on her mathematical expertise, while Mathias and Michael contribute their experience from working on Endpoint Protector.

Implementation proves harder than expected. Yvonne and Mathias have to work closely with the team in Skopje, where decisions often wait for instructions. Different working rhythms and expectations about commitment to the young company slow development and drive up costs. Michael's experience with teams in Southeast Europe cannot simply be transferred either. The founders must find new ways to put responsibility where the work happens.

As financial room for manoeuvre narrows, Yvonne works to secure funding. In mid-2023, her efforts win Laigo approval for the BW Pre-Seed funding programme. It buys the team time. Sadly, following this success, she has to step down as CEO for health reasons.

A Quick Decision and Teamwork

Until then, Michael has mainly financed Laigo and helped guide its strategy. Now he has to act. Within three days, he takes on the CEO role and puts everything else on hold. Holger has his back. What they learned racing together becomes the way the two founders work.

The leadership question is settled for the moment. But the list of urgent tasks keeps growing. The funding must be paid out, the team developed and a potential order from Turkey prepared. Michael brings Benny over from Bauner Consulting as project manager. His role is to connect customers, development and management.

Then, in September, just a month after Yvonne's announcement, Mathias leaves. His departure comes at the worst possible moment. The new structures are not yet established, and the order is drawing closer. With Mathias, Laigo loses a source of both technical and organisational support. The gap is bigger than expected. Much bigger.

While Benny translates the customer's requirements into concrete development work, Michael and Holger have to make up for what is missing. Laigo is running out of time. The order will not wait. Whether the new arrangement can hold is about to be decided.

The First Customer Will Not Wait

The first order comes from a container port where customs documents are processed manually and data entered into logistics software. Each clearance takes at least half an hour. Errors mean more work. The goal is to handle twice as many clearances with the same team, with fewer mistakes.

Laigo learns of the project through a Turkish partner. The task is to capture the relevant information from documents and pass it reliably to the logistics system. This is not a demo; it is the port's daily operation. Michael negotiates, Benny clarifies the workflow with the customer, and the developers build the solution. The Turkish partners sign and cover part of the development costs. Now documents, AI models and logistics software must work together reliably in everyday use.

Infrastructure must not become a bottleneck either. The founders want to control the hardware and data processing themselves, but cannot afford an expensive AI server. When a supplier declines to build the lower-cost alternative, the team sources its own components and develops an affordable AI computer suitable for a data centre. The server is installed with a reputable hosting provider near Lake Constance.

In March 2024, the minimum viable product is ready. In the second quarter, the customs system goes into production. Their own AI models, a protected document workflow and hardware they assembled themselves prove their worth in real use. Laigo has delivered and won its first paying customer. Now it needs to become a dependable business.

The Founders Show What They Are Made Of

In autumn 2024, payments from the Turkish partners stop. Michael travels to see them, but cannot revive the partnership. The customs system works, yet the revenue is missing. A small order from Germany is not enough to support the team.

Laigo has delivered. Even so, Michael and Holger must now preserve years of work and win new customers. At the end of 2024, Raul, a trusted colleague from the CoSoSys years, comes on board. He takes over development planning and strengthens team spirit. Benny brings together the customer requirements. Michael keeps Laigo running with his own money.

A second funding payment buys them time. But new orders and further financial backers do not materialise. Their financial room for manoeuvre keeps shrinking.

Eventually, Michael reduces the development team to three part-time employees and closes the Skopje operation. Benny leaves. Michael can no longer keep Raul on either. By the end of 2025, running costs have fallen to one-sixth of their previous level. The company survives. The remaining team consists of Petrit, Jon and Fuat, alongside Holger and Michael. Together, they call themselves the “Laigorians”.

Still, Michael comes close to losing faith in Laigo. So much of his own money and work has gone into it. Holger has invested too. Both know what the team is capable of technically. The business simply does not yet sustain itself.

Competitive sport has taught them perseverance and the importance of mutual trust. Now Holger gives Michael the support he needs with a few words:

“Giving up is not an option.”

The short sentence pays no bills, but becomes the founders' mantra. Both want to keep going, together.

With three part-time developers, Laigo still has proven components and experience in secure document workflows. Now they must decide what problem the company should solve next, and for whom.

AI That Stays In-House

During a short break, Michael's thoughts keep returning to Laigo. How much longer will the company survive, and what should they build next? Then the idea from the Munich beer garden returns. AI should do more than search documents; it should recognise connections. Sensitive data should stay on site and never leave the company.

A potential niche market provides the starting point. Laigo develops a knowledge archive with semantic search. But it soon becomes clear that the approach has broader potential. Many organisations want to use their knowledge without sending it outside. The question now is how to turn this into a product that can be deployed repeatedly, rather than developed afresh for every customer.

A Turnkey Product

Michael remembers his time at CoSoSys. There, a hardware appliance had brought the breakthrough. It was a complete system, combining software and hardware, that customers could connect and use immediately. “Out of the box.” In March 2026, the two decide to build a local AI appliance. The AI runs on the supplied device, inside the customer's business. It needs no cloud, no internet connection and no data centre. Documents and requests are processed right there on the premises. Confidential information stays in-house and under the customer's control.

The preceding years have laid important foundations. These include modules for documents, search and security, along with experience in affordable AI hardware. The team does not have to start from scratch. It has to bring existing building blocks together into a dependable product.

Built for Reliable Operation

Between the idea and the finished device, a race against time begins. Components, hardware and local processing must work together reliably. By the end of April 2026, Laigo-GPT is ready. The system is capable and stable in operation. The question of how AI could understand documents has become a turnkey local AI product.

Now comes the next test. Can they find the organisations, the customers, for whom this solution is more than technically possible and is urgently needed?

No Half Measures

The temptation is obvious. The product is ready; now they should reach as many customers as possible. But Michael knows this moment from previous product launches. “Everyone could use this” sounds like a large market, yet fails to answer the decisive question. Who needs the solution so urgently that they will act now?

The first requirement has been met. The product is ready. Laigo-GPT runs continuously on the customer's network, even without a cloud connection. More than ten years of experience with Endpoint Protector and CoSoSys's enterprise appliances have gone into its development. Together with Holger, now Laigo's CTO, the team has built a turnkey system for dependable operation around the clock.

Michael follows a proven sequence. First comes a stable product, then an ideal customer profile whose urgent need it fully addresses. Only when that fit is established and the market is large enough does scaling begin.

From April to October 2026, he holds conversations, demos and tests with potential customers across different industries. Some are curious. Others see the benefits but feel no immediate need to act. Interest alone is not enough.

Gradually, a picture emerges of where need and timing align. These are businesses that want to use AI with their internal knowledge, but cannot or do not want to send confidential data outside. For them, the local appliance has tangible value.

Eventually, the first ideal customer profile is clear. The conversations also reveal which industries may be ready next and which need more time. This sequence becomes an important asset for Laigo. The small team knows where to focus its efforts now and where to turn later. Active sales begin in late August and early September 2026.

Customer feedback is positive. The shareholders also see Laigo-GPT as a strong product for that first profile. On the shores of Lake Constance, an idea has become a system that lets businesses work with confidential data without relinquishing control of it.

Michael and Holger take this next step together. They know what they can count on each other for, and what they are working towards.

… to be continued, because giving up is not an option.