Recognizing an old friend

I studied applied math and physics at MIPT, one of Russia's top physics-and-math institutes, and started out in numerical simulation — modeling physical processes, the kind of work that's math all the way down. When I moved into web3 and smart contracts, I didn't expect to carry much of that with me; from the outside it looked mostly like plumbing.

Then a pool full of leveraged positions turned out to be a 2×2 matrix: deleveraging a whole side of a margin protocol collapsed into a couple of coefficients I could update in constant time. Rebalancing a vault across a dozen protocols was a min-cost flow; socializing bad debt across many tokens was a linear map that only looked like it should explode. Finding each one felt less like inventing something than recognizing an old friend — the same math, in the last place I'd expected it.

That's the thread through most of what I do — I look for the structure hiding inside a problem that presents itself as expensive bookkeeping. And I like to carry it the whole way down: derive the mechanism on paper, write the smart contracts, and build the Rust and C# services that run them in production. I've done that across EVM, Solana, Substrate, and Stellar, on protocols that shipped to mainnet and through audits.

That curiosity spills past what has to ship. I taught myself zero-knowledge proof systems the only way I know how — building small ones from scratch, handing a real problem to a SNARK, and writing up what I got right and wrong. Some derivations never make it into production. I write those up too.

Away from the keyboard, the same instinct tends to follow me. I play chess — not master level, but the best I ever have — and one of my earliest projects was a doomed attempt to build an engine for it. I picked up pool a while back for the geometry of it — reading the angles, planning the break, even if my cue still lags my intentions. I also like pointing new tools at everyday problems — lately leaning on AI to sharpen my chess openings and to get back into Korean. I've also been trading films for novels — I used to watch far more than I read, and these days it's the other way round.

What pulls me is a problem with real mathematical teeth — the kind where the shape of the answer isn't obvious until you find the right way to look at it. I'm always glad to talk to people working on those.