I automate manual work with AI agents and ship AI features to production.
Product manager by training, engineer at heart. I start by finding out who will use the product and who will pay for it. Then I build and automate everything myself.
How I can help
Automate a process
LLM agents handle requests, documents, payouts and reconciliations. Everything runs in your infrastructure, and personal data never leaves it.
Examples: EdTech Back Office, Stape Payouts
Describe the processAdd AI to your product
I take your LLM feature or agent to production: quality checks on real conversations (evals), tracing, personal data masking.
Examples: Mental Health App Evals, ERP Agent Demo
Discuss the featureAdvice call
One call about your AI roadmap, architecture or vendor choice. 15 years in software, from QA lead to COO.
Examples: CV
Book a callPath
How I work
Spec interview, plan, AI sub-agents, user acceptance, changelog. I follow the same process on every project, so I ship alone what usually takes a small team.
Stack
TypeScript, Next.js, Node.js, Postgres, Docker. OpenAI and Anthropic APIs, agents, MCP, evals, OpenTelemetry, Whisper.