Current Projects:

1. AI start-up for engineers

Behind the VC: A Personal Journey

Early Interest and the First Fund

I first heard about VC back in my early student years, around 2004. I was struck by the idea that a single successful deal in a portfolio could cover the losses of hundreds of others. It seemed almost unbelievable, so I began studying the field. I was surprised by how much depends on liquidity, and on the belief that market participants place in one another. It felt like a strong psychological argument—much like the invention of the LLC form, which was itself a paradigmatic shift. It is built on the assumption that, in practice, all market participants pursue long-term goals and are aligned by the prospect of future returns. This structure allows startups to test bold—even seemingly insane—ideas, while risk gets diluted at scale.

Many years later, in 2021, a friend of mine called shortly after moving to the United States. By that time, I was already an established scientist, and he asked how financing works in academia. I explained that, unfortunately, many scientific projects die inside laboratories before ever reaching the market.

I gave him an illustrative example from my time in Berlin: a colleague of mine in an optics lab was working on a prototype photonic processor back in 2010. But as the project began drifting toward a commercial product, the head of the lab feared losing government funding for fundamental research. They had to shut down that entire line of development in favor of pure science.

After several discussions, we realized that we had enough complementary expertise to launch our own DeepTech fund. We invited another founder from the financial sector and incorporated a company in Delaware at the end of 2021 (rpv). When the war started in early 2022, various circumstances left the two of us continuing alone.

Naturally, raising capital for a first-time fund was long and painstaking. The fundraising formally ended only by early 2024. But even before that, we had already invested in startups in AI, neuroscience, energy, and additive manufacturing.

It was an exciting role in which I was responsible for sourcing and performing due diligence on new projects. Hundreds of startups passed through my hands—photonics, AI, robotics… Through this process, I gained a deep appreciation for the founder’s personality as a decisive factor. I finally understood why the lab leader in 2010 didn’t take the photonic processor to market: a founder must constantly balance between the possible and the impossible. They must recruit key people, quickly dive into everything from fundraising to marketing, and still remain the intellectual driver of the technological vision.

Trust matters as well. Scientists often look at financiers with skepticism, and in this sense I served as a kind of guarantor—showing that there is a real place for scientists inside VC.

Unfortunately, I was too naïve. Or perhaps too inattentive. From the very beginning, I placed unreasonable trust in my partner. As a result, almost immediately after the fundraising concluded in mid-2024, he attempted to remove me through a weaponized capital call. My friends’ capital was also at stake—people who believed in me and in our vision. I had no choice but to file a lawsuit in Delaware to restore my rights.

This experience taught me countless lessons. I learned how dangerous it is to neglect the legal side of venture work. I had to study Delaware law, hire a legal team, and go through every stage of litigation. By the end of 2025, we had reached the stage where the MTD was denied—an intermediate victory. But the path toward restoring fairness is not yet complete. I promised myself to see it through; otherwise, my assurances to scientists that they had nothing to fear would remain empty declarations.

Turning Toward AI

At the beginning of 2025, I decided to take a break from active VC. I only occasionally advised startups and funds, and shifted my main focus to studying AI. Through my academic connections I had heard about approaches like BERT and GPT ten years earlier, but I genuinely had no idea that this technology would grow out of research groups so quickly. So I began an educational journey to understand the capabilities and limits of these models.

I was pleasantly surprised to discover that the mathematical framework of theoretical physics—down to definitions and terminology—almost perfectly overlaps with the mathematical framework of AI. Many leading scientists in the field turned out to be former theoretical physicists. Half a year was enough for me to grasp the fundamentals of LLMs.

My plan was simple: understand the basics and then approach my acquaintances with a clear proposition—tell me about your workflow, and I will think about how and where it can be optimized with a new tool.

Building a Startup

In May, I went to a friend of mine, the founder and chief engineer of a large consumer-electronics company. Together, we identified huge opportunities—not only for his personal routine but for the broader market. I invited another university friend, who had been working with neural networks for over ten years, and by the summer of 2025 we had built an MVP.

We secured initial angel investments, brought in prodigy-level developers, and began preparing the product. Gradually, this became the main focus of my work. I found myself enjoying the role of a startup CEO. It placed me on the opposite side of the VC table from where I originally entered the industry—and it suits me much better.

Above all, I am building a product together with truly gifted people, creating tangible value in engineering workflows. In some sense, we all feel like Ford’s early team, laying the foundations of an assembly-line approach to engineering.

The typical, logical path in VC is the journey from founder to fund manager. Only then can you truly be helpful—not with abstract advice, but with real skills and real networks. My path happened in reverse, but I am grateful for it.

I have not lost faith in the market. I have not lost faith in equal partnerships that multiply the value of what people build together. On the contrary, I now believe even more strongly in the idea of sincerely wanting to create value for all participants.