I'm a developer who loves building AI developer tools — things that help people write, review, and reason about code faster. Most of my projects sit at the intersection of LLMs and software engineering.
- AI code review & evaluation — tools that automatically judge code quality and catch bugs, using LLMs as reviewers and judges.
- Multi-agent coding systems — orchestrating multiple AI agents (writer / reviewer / tester / refiner) to produce better code.
- Dev workflow automation — commit message generation, code translation, code review bots for PRs.
- GEO / content optimization — helping content get cited by AI search engines (Generative Engine Optimization).
- multi-agent-code-studio — A multi-agent system where an orchestrator coordinates Writer, Reviewer, Tester and Refiner agents to generate and polish code.
- llm-as-judge — Evaluate AI-generated code quality with multiple LLM judges, scoring via ICC consistency check.
- code-eval-platform — A web platform for uploading coding tasks, running cross-model evaluation, and showing live leaderboards.
- geo-optimizer — Analyze and improve content so it gets cited by AI search engines.
- prompt-playground — A zero-backend tool for testing and comparing prompts in the browser.
Python · FastAPI · LLM APIs (OpenAI-compatible) · JavaScript/HTML · Git · GitHub Actions
You'll notice my projects are small and my star counts are modest. I'd rather build things that actually work and are well documented than chase vanity metrics. Every repo above has a README, quick start, and license — feel free to run them.