I will build a production python backend with fastapi or django

Most AI features fail in production because the backend around them wasn't built to handle real traffic, concurrent requests, or scaling load. I build the Python backend layer that supports your product long after the demo works.

Using FastAPI or Django depending on your stack, I set up REST or GraphQL APIs, PostgreSQL data models, Redis caching, and Celery background workers for anything that shouldn't block a request. Deployment is Dockerized and shipped to AWS with a straightforward CI/CD pipeline.

The goal is a backend that's boring in the best way: predictable, observable, and easy for another engineer to pick up. If you're integrating LLM features, I structure the architecture so those calls don't become a bottleneck or a single point of failure.

I'll ask about your current stack, expected load, and integration points before writing code, so the scope matches what you actually need.