How an Idea Became kausable
There was no single eureka moment when research suddenly turned into a business idea for kausable. Instead, Johannes Haux, Dr. Benjamin Herdeanu and Gregor Ramien focused early on a pragmatic question: Could their technological approach become a solution that would also prove itself in the market?
That mindset ultimately led to the creation of kausable. Founded in 2024, the startup aims to translate scientific insights into technology that solves a concrete problem: How can AI systems continue to make reliable decisions even when their environment changes?
To make the transition from research to entrepreneurship, the team received EXIST funding. This gave the founders the necessary space to further develop their technology and build a viable business model from their scientific work.
AI That Learns from Only a Few Examples
This is exactly where kausable comes in. Conventional AI models can quickly reach their limits in dynamic environments: when conditions change, they often need to be adapted or retrained using new data. This costs both time and resources.
kausable therefore follows a “reasoning-first” approach. Rather than relying on text-based language models, the team is developing AI that can understand relationships, learn from only a small amount of new data and respond to changing situations.
At the core of this approach is a so-called world model. Put simply, the aim is for the AI not merely to reproduce familiar patterns, but to understand structures and relationships that it can transfer to new situations. This is particularly relevant wherever systems continuously process sensor and environmental data and need to react flexibly – for example in robotics and industrial processes.
The approach is inspired, among other things, by research on Prior-Data Fitted Networks and the work of Yann LeCun. kausable deliberately positions its technology deep within the software stack: not as another language model, but as a technological foundation for applications that need to make autonomous decisions.
From an EXIST Project to the de:hub Network
As the technology evolved, the company itself also began to take shape. Along the way, kausable made use of various support programmes addressing different stages of the startup journey.
Following its launch with EXIST, an earlier project by the team established the connection to the de:hub Initiative. Today, kausable is part of the de:hub Artificial Intelligence in Karlsruhe. There, the startup has benefited, among other things, from networking with other founders and support in sharpening its business plan.
As the company developed, different building blocks came together: from support for transferring research into business, to consulting and networking, through to access to additional partners.
€12 Million for the Next Stage of Growth
With its latest funding round, kausable has now reached its next milestone: the startup has raised €12 million in seed funding. The round is led by UVC Partners and Entourage, with participation from the High-Tech Gründerfonds (HTGF) and additional investors.
For Johannes Haux, the funding is both a validation and a starting point: it validates the technological approach while creating the conditions to transform what has so far been a highly research-driven team into an organisation with a stronger focus on product development and engineering.
At HTGF, the constructive and founder-friendly exchange with Hendryk Hosemann played an especially important role. The new capital now gives kausable the room it needs to enter its next phase of growth.
From Funding to Internationalisation
Following the seed round, attention is increasingly turning to the next stages of development. For its first steps towards internationalisation, kausable is currently in talks with the German Accelerator.
In the long term, the startup is also looking towards the US market. At the same time, kausable deliberately wants to remain rooted as a German and European company and contribute to strengthening the innovation ecosystem around Heidelberg and Karlsruhe.
This continues the journey from research to company-building and financing in a consistent way: with every new phase of development, the requirements change – and so do the partners and support programmes that become relevant to kausable.
From a Research Approach to a Product
With the new funding, the focus is now shifting above all towards product development. Following an intensive research phase, kausable is specifically strengthening its product development and engineering capabilities. The goal is to build a product layer on top of the technological foundation that can be used and scaled across different applications.
The roadmap for the next three years is clear: continue advancing the technology, build out the product layer and create the conditions for the next financing round – a Series A.
The long-term goal: kausable wants to become the standard building block for applications that need to make autonomous decisions.
A Success Built by the Entire Team
For Johannes Haux, the successful seed round is explicitly not the achievement of the founders alone. He highlights the contribution of the entire research-driven team that has carried kausable to this point.
And that team is now set to grow further. kausable is particularly looking to strengthen its product development and engineering teams. Anyone interested in helping shape the company’s next stage of development can find open positions and further information about the technology at kausable.ai.
Sources:
https://www.htgf.de/kausable-interview/
https://kausable.ai/#home
https://proceedings.mlr.press/v202/nagler23a.html?
https://openreview.net/pdf?id=BZ5a1r-kVsf&utm
https://ai.meta.com/blog/yann-lecun-advances-in-ai-research
