Project Showcases · June 18, 2026 · 1 min read
AI Code Reviewer: I Built a Self-Hostable GitHub PR Reviewer with FastAPI, Redis, and Docker
TL;DR
AI Code Reviewer is an automated system that reviews every GitHub PR. It posts structured feedback on bugs, security, and performance directly in the PR thread. Built with FastAPI, Redis for task queuing, and an advanced AI engine for analysis.
The Problem
Manual code reviews are slow. Developers often wait hours for a teammate to check their code, only to find simple syntax errors or missing edge cases. I wanted a system that handles the repetitive "sanity checks" automatically so humans can focus on high-level architecture.
What I Built
- Automated Feedback: Every PR gets a comment with structured review notes.
- Security Scans: Detects potential vulnerabilities before they merge.
- Performance Analysis: Flags slow code patterns.
- Self-Hostable: Can be deployed via Docker on your own VPS.
Architecture
- Backend: FastAPI (Python)
- Task Queue: Redis + Celery/Worker
- AI: LLM Pipeline
- Deployment: Docker Compose
Hard Parts
Tuning the prompt to be helpful without being annoying. AI can sometimes find "bugs" that aren't actually bugs. I implemented a confidence threshold to only post critical or high-certainty feedback.