Profile
Hands-on engineer and active open-source contributor building AI platforms and developer tools for software engineers, coding agents, and AI-assisted product teams. Turns ambiguous or fragile systems into reliable infrastructure, with experience across agent experience, RAG and hybrid retrieval, MCP, context engineering, automation, ETL, and on-prem AI.
Core expertise
Agent infrastructure · AI product engineering · Developer experience · Agent experience · LLM applications · Agentic systems · RAG and hybrid retrieval · Context engineering · MCP · AI-assisted development · Python · TypeScript · Go · Rust · Bash · Linux · Docker
AI-native development since 2022: Codex, OpenCode, Claude Code, Hermes, and Prime Agent embedded across the full engineering workflow—from coding and review to debugging, DevOps, infrastructure, and systems administration.
Languages: Russian (native) · English (intermediate)
Selected Experience · 2019–Present
Technical Partner & Principal Product Engineer ·
Knowlume
Mar 2026 — Present
- Own the engineering side of an early-stage SaaS product in direct partnership with the founder, from product capabilities and architecture through deployment, infrastructure, integrations, and production support.
- Stabilized critical workflows while reducing some operations from tens of minutes to seconds, and continue to improve the system without interrupting active product development.
Independent AI Platform & Product Engineer · Self-employed
Aug 2025 — Present
- Build internal platforms and agent-native developer tools that help engineers, generalists, and coding agents understand codebases, review changes, manage context, and ship software more reliably.
- Deliver MCP integrations, RAG and retrieval systems, data pipelines, on-prem AI, self-hosted Windmill workflows, and brownfield product recovery for founders and small teams.
Technical Partner · Khabaroff Studio
Nov 2025 — Mar 2026
- Owned technical delivery for private B2B projects: clarified requirements, designed architectures, built integrations, and delivered working systems.
- Stabilized and redesigned an ETL pipeline processing millions of Excel rows: reduced runtime from 20–30 hours to 8 minutes, made results deterministic, and uncovered errors in earlier manual calculations.
Lead ASIC Firmware Engineer → Development Team Lead → Senior Software Engineer · Hiveon
Oct 2019 — Jul 2025
- Joined after contributing improvements to Hiveon’s public projects, moved from technical support into ASIC firmware within two weeks, and developed operational tooling in Go and Bash across heterogeneous ASIC and GPU mining hardware.
- Led a 5–6 person engineering team while remaining a hands-on contributor; later returned to a senior individual-contributor role with staff-level scope, reporting directly to the CTO.
- Reverse-engineered device behavior, reviewed code, resolved production incidents, mentored engineers and QA specialists, and helped teams navigate a complex hardware domain.
- Served concurrently as Forward-Deployed Engineer / Technical Success Partner for Hiveon partners WATTUM (Texas, USA), Adaptic (Sweden), and Block One Technology (Canada), embedding with client teams and turning field failures into product improvements.
Developer tooling, open source & AI projects
Open source: Contributor to ctx.rs, Hermes Agent, Context Gateway, Git. Ship. Done., and OpenClaw.
- Agent-native engineering: Enji Guard CLI/MCP for automated code audits; Cross AI Review for adversarial multi-agent review and planning.
- Coding-agent DX: tools for context optimization, repository analysis, session management, AI-facing documentation, Claude Code workflows, and safe codebase exploration.
- MCP & retrieval: multiple agent-facing MCP servers; dotmd hybrid RAG combining semantic search, BM25, knowledge graphs, RRF, and cross-encoder reranking.