Tech Lead / Staff Engineer based in Amsterdam, working across regulated/safety-critical environments and AI-native engineering — with roots in internet-exchange-scale network automation. CISSP, GCSA, Generative AI Leader.
Architecture: Maintainable systems that still make sense a year from now.
Delivery: Predictable shipping with minimal ceremony.
Operations and security: Infrastructure that scales with the team instead of slowing it down.
GenAI in production: LLM integration, RAG, and observability (Langfuse). Building toward agentic systems and MCP-based tool use.
- Languages: Python, Ruby, Go, TypeScript/JavaScript
- Cloud and infra: AWS, GCP, Kubernetes, Docker, Terraform
- GenAI: LLM integration, RAG, Langfuse
- Security: CISSP, GCSA, Generative AI Leader
- EgressLens − monitor outbound network activity from Python apps running in Docker.
- LLM Observability PoC − a Langfuse-based PoC for LLM observability in Python.
- Exploring: agentic RAG (LangGraph, MCP), LLM evals (promptfoo, Inspect AI), and high-throughput inference (vLLM).
On AI engineering
- STATUS.md: a shared file for multi-agent work
- A minimal LLM Ops stack with tracing and model costs
- RAG: A (mostly) no-buzzword explanation
On security
- Amsterdam, NL
- contact@igortkanov.com
- LinkedIn · Blog
When not writing software, I'm building things on a breadboard. ESP32 firmware, sensors, basic circuits, and the like.