Overview
Deploy a self-hosted Git AI instance for organization-wide metrics
For organization-wide deployment, Git AI offers a self-hosted SCM app that processes webhooks from all your repositories automatically.
When to Use Self-Hosted
Self-hosted deployment is recommended when:
- You need to track 10+ repositories
- You want organization-wide metrics
- You need real-time updates via webhooks
- You want centralized configuration
Architecture
The self-hosted solution consists of:
- Webhook Receiver: Receives events from your SCM (GitHub, GitLab, Bitbucket)
- Worker: Processes git events, and exports metrics to your dashboard
- Database: Stores processing state and cache
All components run on your infrastructure, ensuring your code never leaves your environment.
Deployment Options
Docker Compose (Recommended for Testing)
Quick setup for single-server deployments:
- All components in one docker-compose.yml
- Suitable for up to 100 repositories
- Easy to configure and maintain
Kubernetes
Production-grade deployment for scale:
- Separate scaling for webhook receiver and workers
- High availability and auto-scaling
- Suitable for 100+ repositories
Serverless (AWS Lambda, Cloud Functions)
Event-driven deployment:
- Pay-per-use pricing
- Automatic scaling
- Ideal for variable workloads
Security & Compliance
Data Residency
All code analysis happens on your infrastructure:
- Code never sent to external services
- Full control over data location
- Meets compliance requirements (SOC2, GDPR, HIPAA)
Network Security
- Private network deployment supported
- No inbound connections required
- Works behind corporate firewalls
Get early-access
We've been helping several large teams roll out early versions of the self-hosted exporter. If you're interested in getting early-access set up a call with the maintainers and we can get started / help you plan your rollout: