ML Researcher / AI Engineer — I bring production-infrastructure discipline to LLM agents.
- 🔭 DevOps → ML. Four years as a DevOps engineer: process automation and building an internal DevOps platform used by hundreds of teams. Now applying that systems-and-scale mindset to AI — the infrastructure and reliability of LLM agents.
- 🧠 I build agents that run on models you control, not thin wrappers around a frontier API — memory-as-environment, recursion, and guardrails that make small local models genuinely useful.
- 📍 Based in Moscow.
- ✍️ I write up what I build.
RLM_agent — a stateful agent on a Recursive Language Model engine. A local LLM drives an IPython REPL, queries sub-models from its own code, and consolidates a durable memory after every episode. A study in getting real work out of a small, local model through structure rather than scale.
Languages
ML / AI
Infra / DevOps
Backend
Links coming soon.