Systems Engineer · Platform Builder
From first engineer to leading a product used by thousands daily.
Now building the infrastructure layer that makes agents ship code — not just suggest it.
Proof of Work
HR-tech startup → Kaspersky acquisition
Problem
Early-stage product with no engineering team, no architecture, no processes.
Solution
Joined as first engineer. Built the team, established architecture, led Vue 2→3 migration without stopping delivery.
Result
Product scaled to thousands of daily users across major enterprise clients. Releases went from stressful events to routine deploys.
KB Labs — solo-built platform
Problem
AI agents generate code, but nobody controls how. No observability, no policy enforcement, no reproducibility.
Solution
Designed and built a 125-package platform: plugin system, workflow engine, semantic search, agent orchestration.
Result
Full engineering automation stack — self-hosted, policy-first, built by one engineer. OSS core shipping today.
Scaling engineering processes
Problem
Manual code reviews bottleneck delivery. Quality depends on who reviews, not on what's reviewed.
Solution
Built AI review pipeline with heuristic + LLM layers, baseline ratcheting, regression detection.
Result
Automated quality gates that catch regressions before merge. Review quality became consistent and measurable.
About
Most teams talk about AI-assisted development. I built the platform. 125 packages, 18 monorepos, agents that actually ship code — not just autocomplete it.
First engineer at a Kaspersky subsidiary → team lead → architectural owner. Now I sit at round tables with C-level discussing how to make agent-driven engineering predictable at scale — and then go home and build it.
We switched to tractors, but the field stayed the same size. I'm working on the field.
Technologies I work with
Experience
Open-source platform for engineering automation. Plugin system, workflow engine, semantic search, agent orchestration — OSS core with commercial layer ahead.
Joined as the first frontend engineer at an early-stage HR-tech product inside Kaspersky. Grew it from startup to acquisition — used daily by thousands of employees at major enterprise clients.
Co-founded a web studio. Learned that the hardest part of engineering isn't code — it's translating business needs into technical decisions under real constraints.
Where it started. Corporate sites, landing pages, learning to write code that other people have to maintain.
Philosophy
Observability
If you can't see it, you can't control it
Agent Predictability
AI output you can verify and trust
Cost of Coordination
More features, same bottleneck — unless you fix it
Time to Confidence
How fast can you trust a release