About
AI-Native engineer. Tech lead. 11 years on Toptal.
Background
I started in physics infrastructure — junior scientist and web developer at the Budker Institute of Nuclear Physics in Novosibirsk, building configuration databases and integration tooling for experimental hardware. In 2009 I went to CERN for three months to work on the LHC accelerator software stack. I learned Python in two days and shipped a working solution — connection pooling, multi-process parallelization — in two weeks; the rest of the stay went into tests and documentation.
I joined Toptal in 2015 and spent the next eight years on their network — first as a full-stack developer for client projects across health insurance, computer vision, and edtech, then as the tech lead on Toptal's own conversion engine. After I rotated off, the team handed it to a 10-person group that continued ownership.
Since 2024 the work has shifted toward AI: nationwide legal data pipelines for FirmPilot, multi-agent orchestration for my own products, and architecture review for SaaS founders preparing to scale. I work from Almaty, with clients across the US, UK, and EU through Toptal.
How I work
AI-native by default — not retrofitted onto an existing process. Most of my engineering now runs through Claude Code with a custom orchestration setup I built and refined over a year of daily use. Specifications, code, tests, documentation, and review cycles all flow through the same pipeline. Colleagues use the setup as a reference for their own configurations.
The output stays production-grade: tests, observability, handoff documentation, and small reusable libraries that the rest of the team can adopt without me in the loop. When I built the FirmPilot court-opinion parsers I shared the proxy and tooling with the team so the next state-level parser took hours instead of days.
I prefer to lead small teams — two to four people — where every decision compounds and there's no organizational drag between idea and shipped result. R&D pace, production discipline.
Beyond engineering
For three years (2011–2014) I ran the R&D department at ASIA Consulting Group — later OneBox, eventually acquired by Ernst & Young. I was the only engineer in a team of management consultants, building the technical side of a Gallup-style company-diagnostic algorithm: a deep questionnaire that produced a structural model of the client organization. I also sat in on strategy sessions, HR assessments, and executive workshops as the technical voice in the room.
I hold a postgraduate degree from the Budker Institute, master's and bachelor's degrees from Novosibirsk State University, and a diploma from Russia's Presidential Management Program (2010) — formal training in finance, organizational strategy, and management. Earlier I went through Intel's School of Leaders for Innovative IT Projects at NSU (2009).
The reason this matters in practice: I can sit between business stakeholders and engineers without losing precision in either direction. Boards get a clear technical answer; engineers get a clear business reason. That bridge is rare in senior engineers and is where I'm sharpest.
Tell me about your project
I take on a small number of engagements at a time. If you're building something where engineering judgement and AI-native execution would compound, get in touch.
Work with me through ToptalToptal handles contracts and payment. Or book a 30-min intro call