Senior AI Engineer · Go & Python · LLM, RAG, MCP, AI Agents · High-load distributed systems
I build production AI systems on top of high-load Go backends: multi-agent workflows, GraphRAG, MCP servers and AI-assisted engineering pipelines — at scale of billions of requests per second.
- 💼 Senior AI Engineer | Senior Go Engineer — observability platform, 2B RPS, 30K servers, 4 PB of data
- 🤖 Building LLM agents, GraphRAG and MCP integrations on top of production infrastructure
- 🛠️ Agentic development daily: Claude Code + custom MCP tools to ship and maintain 6 client SDKs
- 🌱 Open-source contributor to StatsHouse — 70+ merged PRs
- 🧠 Background: search engines, recommendation systems, event-driven microservices
- 👨🏫 Taught Go & algorithms to 150 students at ITMO University
- 🎓 ITMO University — Applied Mathematics & Computer Science
- 📍 Frankfurt, Germany · open to remote
- 🔎 GraphRAG pipeline — Neo4j + pgvector hybrid retrieval, LangGraph orchestration; powers a multi-agent framework for 200+ engineers, 88% accuracy on a 300-question eval set
- 🧑⚖️ LLM code review system — adopted by 12 teams, ~130 PRs/month, 92% of comments accepted, review time −20%
- 🔌 MCP server for observability — AI agents query metrics, assess system health and explain incidents in natural language; incident resolution time −15%
- 🧭 AI semantic search over metrics in StatsHouse via MCP tool calling
- ⚡ Agentic coding at scale — single-handedly maintaining SDKs in Go, C++, Python, Java, PHP and Rust with Claude Code and custom MCP tooling
LLMs & Model Providers
Agents & Protocols
RAG, Vector & Graph Search
Agentic Coding
Evals & LLMOps
Languages
Frameworks
Databases
Message Brokers
Protocols
Cloud & Infrastructure
Observability & DevOps
- 🤖 AI agents & multi-agent systems — tool calling, ReAct, human-in-the-loop, LangGraph workflows
- 🔌 MCP servers in Go — exposing production systems (metrics, logs, search) to LLM agents
- 🔎 RAG & GraphRAG — hybrid vector + graph retrieval, embeddings, eval-driven quality
- ⚙️ Agentic software development — Claude Code, custom MCP tools, AI code review in CI
- 🏗️ Microservices & distributed systems — event-driven architecture, Kafka, transactional outbox
- 📈 High-load platforms — 2B RPS, rate limiting, shard isolation, controlled degradation
- 🔍 Search & recommendations — Elasticsearch, hot/cold storage, ML-model serving