AI Research Engineer
My ongoing interest is devoted to building systems with the capacity to self-improve. Technology must be alive and communicable. Wide expertise from 10+ years across science & industry. Let's chat.
Drop me a message — if you'd like to work together, or you're just curious.
About
I'm an AI Research Engineer with a builder's attitude and broad expertise from 10+ years of experience. My ongoing interest is devoted to building systems with the capacity to learn and self-improve.
Since 2016 I've worked in R&D across academia and industry — a university lab, a couple of hardcore tech companies, and now a research-centric startup pursuing next-gen AI that combines category theory, logical solvers, and agentic reasoning. My knowledge spans formal Computer Science (programming languages, translators, type systems), classic ML (you name it), and modern agentic AI. I have several publications from interdisciplinary scientific collaborations, plus ongoing side projects.
I've also had a period of contemporary-art practice — several acclaimed interactive-media works and a city-scale exhibition I co-curated, with 3000+ visitors and wide media coverage. That chapter shaped my creative approach to exploring and testing ideas.
I travel a lot. In my spare time I study classical Sanskrit, practice Vipassana meditation, and prototype new product and research ideas.
Looking deeper than the present — and building for the future.
Interested in collaborating on
Today's agentic systems are largely ad-hoc engineering. LLMs are fuzzy, so building reliable agents — and multi-agent systems — is genuinely hard, and the research is only catching up. Much of it rediscovers old lessons from distributed computing and cybernetics. It's the right moment to build real foundations for scalable agentic systems.
Custom DSLs and the evolution of languages. A task-specific language can be surprisingly effective — it carves out a much narrower search space. Most of my work rotates around this, and I've seen its power firsthand: a frontier joining formal methods with the strength of LLMs, while staying close to applications — from biology to legal to anything else.
Neurosymbolic AI, evolutionary AI, novel architectures. AI waves supersede and reinforce one another — RL applied to Transformers is what brought us agentic systems. That pattern will repeat, and there's a trove of techniques and insights still to be combined. Excellent research territory.
Curriculum Vitae
I learn and use whatever is required to achieve the goal. In the AI age, any problem is approachable.
AI & Reasoning
Formal Methods
Mathematics
Languages
ML & Data
Web & Product
Infra & Craft
Teaching & public-lecture experience · academic English (C2) · creative & product design.
Reasoning-AI startup · Tokyo
Senior Research Engineer
High-quality R&D across the company's core projects — in close collaboration with researchers and engineers — at a startup building next-gen reasoning AI from category theory, logical solvers, and agentic methods.
AI in Industry Lab · Saint Petersburg
AI Research Engineer
Leading development of GOLEM — an AI framework for graph optimization with metaheuristic methods (genetic, swarm). Driving architecture, research direction, and a small research team.
V.V. Dokuchaev Soil Museum · Saint Petersburg
Exhibition Curator
Co-curated an international exhibition of technological art dedicated to soils — 28 artists, 20 art objects and installations across 2 venues, 3000+ visitors, ~100 media publications, and a parallel educational program of 10+ events.
Saint Petersburg
MA in Arts, Art & Science (with Honours)
A master's program at the intersection of contemporary art, science, and technology — the ground for the multidisciplinary lens I bring to research.
MPS — Meta Programming System
Software Engineer
R&D on Coderules, a Prolog-like meta-language for declaratively defining type systems inside MPS, the IDE for domain-specific languages.
jetbrains.com/mps mps-coderules
Talk — “Coderules, a new typechecking engine”, MPS Users Conference, Amsterdam, Oct 2019.
Bachelor Thesis
Researcher
Developed an embedded C-like DSL and a library for programming small-scale heterogeneous microcontroller systems, working with LLVM IR and dynamic code generation.
Software Intern
NLP Engineer (Intern)
Built an NLP framework for extracting structured information for marketing research — entity extraction, web ontologies, and RNNs.
Data Science Intern
Data Scientist (Intern)
Improved industrial SSD caching algorithms with predictive machine learning — a research article followed by product integration.
Software Engineering
BSc in Computer Science (with Honours)
Bachelor's in Computer Science with a strong foundation in formal systems, algorithms, and software engineering.
Publications & Talks
Coderules, a new typechecking engine
Projects
Professional
Senior Research Engineer
A startup building next-gen reasoning AI from category theory, logical solvers, and agentic methods. I designed and built the bidirectional bridge between the in-house formal DSL and SMT solvers — including formal translation of a coroutine-like language, foundational for the system's reasoning — and ran R&D on solver coverage and speed, ARC-AGI approaches, and subgraph matching.
Graph optimization & learning
A framework for graph optimization and learning by adaptive, hybrid evolutionary algorithms. Built around two ideas — universality of the approach and adaptability of the algorithm — and used by several teams at the ITMO AI Institute.
Automated modeling
An AutoML framework for the automated modeling of composite pipelines, with GOLEM serving as its evolutionary optimization core.
Type systems for DSLs
A declarative, Prolog-like engine for defining the type systems of domain-specific languages in JetBrains MPS — including a prototype for incremental, language-independent type checking.
Personal
Attention-aware timer · Web & iOS
The first productivity timer that's aware of you and adapts to you — it follows your focus, pulls you back from distractions, and surfaces insights (and, I'd argue, it's the most aesthetic timer out there). Built on a research-informed attention-tracking state machine with custom gaze-tracking and brush-drawing. Solo, end-to-end: idea, research, engineering, product, landing, and go-to-market.
Contract-learning for ARC-AGI
A neurosymbolic ARC-AGI solver that wraps Poetiq's record-setting refinement loop in a machine-checkable layer: a typed DSL paired with z3 contracts over an abstract grid domain, sound by construction. Candidate programs are refuted before they ever run — 99.4% refutation power, zero false refutations across 13,641 checks — and an LLM-free synthesizer grows its own DSL to solve tasks with no model call at all. Verified invariants cut the median iterations-to-solve from 8 to 3, at zero harm.
Crypto market microstructure
An autonomous R&D program that mines crypto-market microstructure for tradeable edges: a real-time arbitrage-graph engine feeding an agentic research loop that tests falsifiable hypotheses against live cross-market data and graduates the survivors toward execution (testnet → live). The founding premise — risk-free negative cycles — proved productively wrong (all stale-quote artifacts), and the meta-finding is sharp: every retail-scale micro-edge is real but sub-fee. One survivor cleared the bar — the tokenized-gold basis, a market-neutral spread uncorrelated with BTC, gold, and USDT stress.
DIE · artificial life
An artificial-life project reproducing the emergence of distributed intelligence under environmental pressures, using neural cellular automata grounded in Evolution, Reinforcement Learning, and Active Inference.