Deployment

Production deployment guides for LLM Proxy.

What’s in this section

For production deployments, we recommend AWS ECS with CDK:

  • Aurora PostgreSQL Serverless v2 for database
  • ElastiCache Redis for caching and rate limiting
  • ALB with ACM for HTTPS termination
  • Auto-scaling based on CPU/request count
  • ~$130/month for low-traffic deployments

See the AWS ECS Architecture Guide for details.

Alternative Deployments

  • Docker Compose - Good for development and testing
  • Kubernetes/Helm - For existing K8s infrastructure (see #49)

Table of contents