🚀 Inspiration
Modern software teams struggle with manual code reviews, missed security vulnerabilities, and lack of proper test coverage. We wanted to automate this process using AI to make development faster, safer, and more efficient.
💡 What it does
CodeGuard AI is an intelligent code review agent that integrates with GitLab and automatically analyzes merge requests. It performs:
- 🔍 Security analysis (detects vulnerabilities like SQL injection)
- 🧠 Code quality review (suggests improvements)
- 🧪 Test case generation (auto-generates test scenarios)
It also provides a UI dashboard where users can manually test and analyze code snippets.
🛠 How we built it
- Backend: FastAPI
- AI Engine: OpenAI (GPT-based analysis)
- Integration: GitLab Webhooks & API
- Deployment: Render
- UI: Custom HTML/CSS dashboard
The system listens to merge request events, processes code diffs using AI, and automatically posts feedback as comments on GitLab.
⚡ Challenges we ran into
- Handling async processing without webhook timeouts
- Parsing structured JSON responses from AI reliably
- Deployment issues with port binding and environment variables
- Designing a UI that clearly explains AI outputs
🏆 Accomplishments that we're proud of
- Fully working end-to-end AI code review system
- Real-time GitLab integration with automated feedback
- Clean and functional UI for demo and testing
- Successfully deployed and publicly accessible
📚 What we learned
- Building scalable AI-powered backend systems
- Handling real-world API integrations (GitLab + OpenAI)
- Deployment debugging (Docker, Render, Railway)
- Designing user-focused developer tools
🔮 What's next for CodeGuard AI
- Support for multiple repositories and teams
- Advanced vulnerability detection (OWASP-based)
- CI/CD pipeline integration
- Code suggestions and auto-fixes
- Team dashboard with analytics
Built With
- css
- fastapi
- gitlab-api
- html
- javascript
- openai-api
- python
- render
Log in or sign up for Devpost to join the conversation.