Kyrylo Kalashnikov on Leaving Ukraine at 13, Judging Frontier Work, and Building Technology That Could Change Science

Kyrylo Kalashnikov has built an unusual career unusually fast. He is an engineer and founder whose work stretches across artificial intelligence, robotics, instrumentation, and biotechnology. He has been a two-time Emergent Ventures grantee, a New Science Fellow, a 1517 Medici grantee, a peer reviewer for academic journals, and a judge for multiple global AI hackathons with prize pools exceeding $100k and thousands of participants around the globe.

At Neuralink, he was a software engineer on the surgical robotics platform behind the first FDA-approved brain-computer interface implantation procedures in history, and responsible for the foundational architecture of their next-generation system. He is also the builder of an open-source $500 self-driving lab now integrated into the University of Toronto's graduate curriculum and recognized by the Acceleration Consortium, the world's leading self-driving laboratory research initiative. In this conversation, he talks about leaving Ukraine alone at 13, learning by building before he felt ready, and why he believes the most consequential engineering work is moving into the physical world.

You left Ukraine for North America alone at 13. How did that shape the way you work now?

It made self-direction non-negotiable. I was not in a position to wait for the perfect support system or the perfect curriculum to appear around me. I had to adapt.

I moved to another country at an early age with limited English. I learned that catching up is an active process. There was no option for me to sit back and wait for the gap to close on its own. That mindset carried into everything else. If I wanted to learn something and it was not available locally, I would find the textbook, build the prototype, and keep pushing until I understood it.

A lot of people your age are still trying to figure out one field. You have already worked across machine learning, robotics, hardware, and biology. Why work that broadly?

Because the problems I care about do not stay inside one discipline. If you are interested in consciousness, aging, scientific discovery, or advanced instrumentation, you quickly run into boundaries. Machine learning alone is not enough. Biology alone is not enough. Hardware alone is not enough.

I never found the idea of staying inside one narrow technical box very convincing. The more serious the problem, the more likely you need to think across systems. That is where much of the real leverage lies.

Your work on the open-source self-driving lab got a lot of attention. Why did that project matter so much?

I had already spent time working with robotic laboratory systems that cost hundreds of thousands of dollars, sometimes millions. And I kept noticing that what those machines enabled was not fundamentally more sophisticated than what cheaper components could theoretically do. The cost was not tracking the capability.

So I started reasoning from first principles. If you strip away those assumptions and just ask what the system actually needs to do. For instance, move a liquid, apply a voltage, measure a response, feed that back into the next experiment. You find that consumer-grade robotics, open microcontrollers, and a few smart software choices can get you most of the way there. The result was a fully autonomous electrochemistry platform for about $500 in hardware.

That matters because access shapes who does science. If only elite institutions can afford automation, the field develops through a bottleneck. I wanted to build something serious enough to be useful and cheap enough to spread. Since then, it is being incorporated into a graduate course on self-driving laboratories, been recognized by other labs, and is fully open to replicate from an open-source repository.

You were involved in early-stage development of the next-generation surgical robotics at Neuralink during your tenure. What did that experience teach you?

It taught me that reality strips away abstraction very fast.

When you are working on a surgical robotics platform intended to help scale brain-computer interface implantation, there is no room for fuzzy thinking. The system has to work. Every layer matters. Software, hardware, architecture, reliability, all of it becomes concrete very quickly.

It also reinforced that high-stakes engineering rewards people who can move between layers. You cannot hide inside one specialty forever when the system itself is the product.

You have also been recognized through grants and fellowships. What do those selections mean to you?

Honestly, less than the work itself, but they are useful signals in a specific way. When you are working on problems that do not fit neatly into an existing institution or career track, external validation from people with good judgment helps establish that the direction is credible.

Emergent Ventures backed me twice, in 2022 and 2025. The 1517 Medici grant was the first outside bet on my micro-robotics work, which at that point I was literally developing computer chips in my bedroom. The New Science fellowship introduced me to some of the most high-agency researchers who are genuinely trying to move the frontier of science.

What those selections share is that they come from people who are hard to impress with credentials alone. That is the part that means something to me.

How do you think about giving back to the communities you have been part of?

I think it is very important to give back to the community, since at the end the work I am doing is for other people.

The most direct is open-sourcing the work. The self-driving lab being replicated by different institutions throughout the world.

Writing is another. The Substack started as a way to think in public, but it became a channel for connecting with people who are working on similar problems. Some of the most useful conversations I have had came from people who found the writing first.

From judging hackathons to peer reviewing papers, I have also tried to contribute on the evaluation side of the field. It all forces you to make your standards explicit in a way that building alone does not always require. What counts as genuine novelty? What is technically impressive but goes nowhere? That kind of thinking sharpens your own work as much as it helps others.

So what does that look like in practice? How have you actually contributed to those fields?

The through-line is instruments and systems that extend what science can do.

At the University of Toronto I built a fully autonomous chemistry robot for about $500 that can run scientific experiments without human intervention.

At Neuralink, I wrote software deployed in FDA-approved human surgeries, and separately built the foundational architecture of their next-generation surgical robotics platform

At Synelligence, I am developing label-free Raman spectroscopy tools for single-cell drug discovery.

At Aion Bio, I am building closed-loop platforms that use electromagnetic and acoustic fields to read and write bioelectric signals in deep tissue.

You mention you write publicly on Substack. Why does that matter alongside the engineering work?

Because public thinking shapes my technical work by creating a fast feedback loop on my ideas

When I am writing about a problem, I am still inside the question. It forces a kind of clarity that talking to colleagues does not always produce. You cannot hide behind shared assumptions when you are writing for people who do not share your context.

The Substack has grown to thousands of readers and paid subscribers, and it became one of the reasons Alexey Guzey identified me and invited me to apply to the New Science cohort. It has also connected me with researchers working on adjacent problems I would not have found otherwise.

What keeps pulling you forward now?

Two big problems. Aging and consciousness.

Those are the questions that keep reappearing no matter what project I am on. They are large enough that I do not expect quick answers. But they are also consequential enough that they justify a long arc of work.

The older I get, the more interested I am in building tools and systems that actually widen what science can do.