‍

Kaya Tilev has joined CV VC as General Partner. A mathematician by training, a founder by experience, and an investor by conviction, Kaya brings more than a decade of experience at the centre of the AI ecosystem: from the Stanford AI Lab to Microsoft's startup practice in North America, to CV VC's mission of backing the AI native future.

We sat down with Kaya to discuss his journey from builder to investor, how he evaluates founders, and why he believes Zurich may be the best place in Europe to build in deep tech.

‍

Kaya, tell us about yourself.

I was born in Kentucky and grew up in Seattle. I studied theoretical mathematics and computer science, then spent a decade in AI in New York, where I managed Microsoft's startup practice in North America. I also spent time at Stanford, at the business school and at the AI Lab, where I published three papers.

I started my career as a quant on Wall Street, but at some point I realized that I wanted to build. I went on to found companies, including Occam's Razor, an early AI company out of Stanford that processed the entire body of scientific literature on Parkinson's and Alzheimer's to search for therapeutic solutions. Difficult problems seem to attract me.

As a founder, I have worn the CTO hat, the CFO hat, the product hat. It is never roses and green fields. So as a GP at CV VC, I bring genuine empathy for founders, understanding the difficulties & hurdles they often face. We’re not just a check. We are builders with a technical foundation, something you don’t find all that often in Europe.

‍

When did you realise AI would define your path?

There was a defining moment in 2016, when I acquired the Twitter handle AI Ventures. I still have it. By building with the technology, you could see the power was there. It was like an uncalibrated engine: it needed more work, but if you spent enough time with it, you knew what was coming. From that point on, everything I built over the last decade had AI in it, one way or another. And I believe this is the technology of the next decade.

‍

A lot of venture money is flowing into AI. Some founders simply add an AI component to attract capital. How easy is it to see through that?

It is pretty easy. We focus on early-stage deep tech, so I like to see at least one founder who has published a paper, spent time in a lab, or contributed to open source in a serious capacity. Or someone who is deeply self-taught; curiosity is king.

I call those other projects bolt-ons. You have a coffee cup, and now you have a coffee cup with AI. When AI is bolted on rather than built bottom-up with the technology, most VCs will see through it. My recommendation to founders: treat artificial intelligence as a tool in the tool chest. It is about how you use it, not attaching "AI" to everything and hoping it sells.

‍

CV VC recently backed Airwayz and Rings AI. What convinced you?

Airwayz is air traffic control for the zero to ten thousand foot drone airspace, a very serious problem right now. The founder is a former military air traffic controller: incredible domain expertise on this exact problem, combined with the willingness to build a company around it. That is rare, and it makes the field defensible. The technology is actively deployed today, protecting sovereign airspace, securing the port of Rotterdam, and has been implemented to monitor stadium airspace at the recent football World Cup.

Rings AI are different. Martin, the founder, is a former venture capitalist who has been building and perfecting the product for seven years. The CRM segment is hard; Salesforce has it under lockdown for a reason. What shocked me about Rings is that it has all the features. And it is AI native in the true sense: built bottom-up on a graph database, the same architecture that powers the large social networks. For any relationship-centric organisation, that is incredibly powerful.

‍

Where do AI and Blockchain converge?

Payments remain the original promise of Blockchain, and that promise is not yet fully realised. Within that category, fraud and compliance are interesting: once you have payments, you are dealing with all sorts of actors, and any globally adopted financial service will have to meet the standards that sovereign states require.

On the agentic economy transacting natively in crypto, I am more cautious. Breakthroughs must happen first: Clear legal frameworks across multiple countries, broad institutional adoption, and banks are not fast institutions. It will happen, but at global scale, touching more than a billion lives, it will take time. That said, if someone built a novel solution where the transaction and the KYC are sandwiched together, that would be a genuine technological breakthrough.

‍

How do you evaluate founders?

I look at three things.

First, deep domain expertise. Are they a genuine expert in their field? For AI founders specifically, I want to see that expertise rooted in the technology itself, not just applied on top of it. I like to see at least one founder who has published a paper, spent time in a research lab, or contributed seriously to an open source project. Or, failing that, someone who is deeply self-taught and genuinely curious about how the models work under the hood. Curiosity is king.

That distinction matters because it separates genuine AI companies from bolt-ons: a coffee cup that becomes a coffee cup with AI stapled on. When AI is bolted on rather than built bottom-up with the technology, it is usually easy to see through, and most VCs will. Artificial intelligence is a tool in the tool chest. It is about how you use it, not attaching "AI" to everything and hoping it sells.

Second, the early category dent. Can their accumulated knowledge make an early dent in the category? It does not need to dominate it, just a dent, a wedge. Once you have a wedge, it means the product is doing something novel that is genuinely attractive to customers.

Third, sellability. Can this product be sold in its early, somewhat scrappy form? If the product is exciting enough, it usually can.

‍

How do you look at Zurich as a location?

Zurich, and specifically ETH (Swiss Federal Institute of Technology Zurich), is possibly the best location in Europe for AI, deep tech, and robotics. Hard things. And hard things are what move the needle. Outside of Silicon Valley, if I had one choice of where to be, it would be here.

The magic trick in the Valley is "yes, and." There are always plenty of reasons why something will not work, but you can also just try, intelligently. What actually makes the Valley work is the ecosystem, and that ecosystem exists here in Zurich too. Our job is to champion it and make it better. Sometimes the Swiss doubt they can operate at the level of the Bay Area. Here, I think we really can.

‍

Looking three years ahead, how will the world look?

With AI, the cycles turn faster because the technology is self-reinforcing. The next three years will feel like five. I do not think it will be doomsday; humanity's edge is collaboration. What I expect is change in the physical world: novel materials, novel alloys, machinery with designs humans could not have produced. When AI crosses into the physical world, that is when we will all know it is real.

‍

With Kaya joining as General Partner, CV VC strengthens its capability at the core of its investment thesis: backing founders building the AI native future across digital finance, automation, and intelligent B2B. Welcome to the team, Kaya!