AI-native developer tools / agentic QA / platform systems
Senior Software Engineer building AI-native developer tools, agentic QA systems, and full-stack platforms.
I build tools around AI agents, developer productivity, frontend/platform engineering, test automation, and human-in-the-loop verification.
Featured projects
Practical experiments in UI QA, source discovery, runtime evidence, engineering memory, and the next layer of developer tooling.
Open source, active build
An open-source AI/devtool project for agentic UI QA and developer tooling.
Problem
Modern coding agents can generate interfaces quickly, but they still need runtime evidence, calibrated critique, repair loops, and human verification to ship reliable UI changes.
Stack
TypeScript / Playwright / LangGraph / RAG / LLM critique / Screenshots / DOM evidence
Open source, early exploration
An exploration of turning Git history into engineering signal for coding agents.
Problem
Useful engineering habits are trapped in commits, diffs, and review history. Compactor experiments with extracting those patterns into reusable signal.
Stack
Git analysis / Diff mining / Signal extraction / Pattern mining / Agent skill generation
Recent notes
Essays and operating notes on AI agents, verification, engineering systems, and the parts of developer tools that have to earn trust.
A short note on why the next wave of developer tools should be built around agents, evidence, and human judgment.
Generating code is only one step. The product is the loop around understanding, verification, repair, and trust.
Why serious coding agents need evidence handoff, calibrated uncertainty, and human review points.