FastAPI

REST is predictable and universal. Every LLM provider, vector DB, and backend you build follows these conventions.

REST is a library catalog — each book has a unique ID, and everyone uses the same rules to browse, add, or remove.

Scenario

You need an API where users upload documents and ask questions about them.

Solution

POST /documents uploads a file, GET /documents/{id} returns metadata, POST /chat returns an answer with source citations.

Do this before moving to the next module — reading alone is not enough.

On paper or in a note, design 4 REST endpoints for a simple notes app: list notes, get one note, create a note, delete a note. For each, write the HTTP method, URL path, and what JSON it sends or returns.

Highlighted lines show where FastAPI happens in the code.

1# FastAPI — minimal example2from openai import OpenAI3
4client = OpenAI()  # create API client5
6# Ask the model to explain this topic7response = client.chat.completions.create(  # core API call for FastAPI8    model="gpt-4o-mini",9    messages=[10        {"role": "system", "content": "You explain fastapi clearly."},11        {"role": "user", "content": f"What is fastapi?"},12    ],13    temperature=0,14)15print(response.choices[0].message.content)  # show output for debugging

• Using GET to modify data
• Verbs in URLs like /getUser
• Returning 200 with errors in body

Quick recap — the most important points from this module.