2025

AI Code Review Platform

Multi-LLM automated pull request analysis

Built to scratch my own itch — manually reviewing PRs at scale is slow and inconsistent. The platform ingests a PR diff, passes it through multiple LLM providers in parallel, and synthesizes a structured review with severity-ranked findings. Uses Groq for speed and Gemini for depth on larger diffs.

TypeScriptNode.jsGroq APIGemini APINext.js