All use casesUse Case
Build a scalable document ingestion and extraction API.
Accept file uploads, process them asynchronously via background workers, extract structured data with LLMs, and store results — all in a battle-tested FastAPI pipeline.
FastAPICeleryRedisOpenAIPostgreSQLPydantic v2
The usual pain points
- ✕Handling large file uploads without blocking API threads
- ✕Processing documents asynchronously at scale
- ✕Extracting structured data from unstructured documents
- ✕Storing and querying processed results efficiently
How the kit solves them
- Async file upload handling with streaming to object storage
- Celery worker pool for parallel document processing
- LLM extraction pipeline with structured output (Pydantic v2 schemas)
- Postgres with JSONB columns for flexible structured data storage
Example implementation
@router.post("/v1/documents/process")
async def process_document(
file: UploadFile,
key: APIKey = Depends(get_api_key),
):
doc_id = await storage.upload(file)
job = await extraction_queue.enqueue(
extract_document,
doc_id=doc_id,
schema=ExtractedInvoice,
)
return {"doc_id": doc_id, "job_id": job.id}
@celery.task
async def extract_document(doc_id: str, schema: type):
text = await storage.read_text(doc_id)
result = await llm.extract(text, output_schema=schema)
await db.save(doc_id, result.model_dump())Ready to build your document processing pipeline?
FastAPI AI Kit ships with everything shown above, pre-configured and production-ready. Clone the repo and start building in minutes.
Ready to ship your AI backend this weekend?
Join developers who skipped weeks of boilerplate and went straight to building.
No subscriptions · One-time payment · Lifetime updates