Full-stack engineer using AI as a tool: TypeScript/Python fixes, automation, CI triage, docs, and careful PR review.
I build small, verifiable fixes for open-source projects and developer workflows. The work I like most is practical: reproduce the issue, keep the diff narrow, run the checks, and leave reviewers with clear validation notes.
- CI Failure Triage - 24h Developer Debug Note: send one failed CI log or PR link and I will return a concise debugging note with likely cause, suspicious files or commands, one recommended fix, and a verification step.
- Best fit: GitHub Actions, TypeScript, Python, lint, test, or build failures where a focused second pair of eyes would save time.
- Scope: diagnosis and written guidance only. No secrets, private credentials, deployment, broad rewrites, or security testing.
- Full-stack TypeScript/React and Python tooling
- GitHub Actions, CI triage, test repair, and release hygiene
- Automation that removes manual steps without hiding the reasoning
- Open-source maintenance PRs with focused scope and reproducible verification
I use AI as a pair-programming and research accelerator, not as a substitute for review. Before I submit work, I check the scope, read the code path, run the relevant tests, and keep sensitive data out of public repos.
- Memanto examples: LangGraph memory flow and Claude Code skills bridge
- Dokploy template validation
- GSM evaluator parsing fix
- Requestly UI/devtools exploration
TypeScript, React, Node.js, Python, GitHub Actions, Playwright, Vite, docs, CI, automation.
If you want a small bug fixed, a failing check explained, or a PR made easier to review, that is the kind of work I enjoy.