All stacks
/// Backend · AI-friendly server
FastAPI services for AI-heavy products.
When the back-end is half AI pipeline, half API, FastAPI is our default. Async Python, pydantic typing, OpenAPI docs out of the box.
/// What we ship with Python / FastAPI
How we use it
in production.
- FastAPI + pydantic v2 + async/await
- LLM integrations: OpenAI, Anthropic, Google
- Vector stores (pgvector, Pinecone, Weaviate)
- Background workers with Celery / RQ
- Postgres / Redis / S3
- OpenTelemetry, OpenAPI, Sentry
/// Python / FastAPI in production
