Python API made with AI: moving FastAPI or Django from development to production

Published on 2026-10-27 by frfrand

How to deploy FastAPI and Django created with AI: ASGI, WSGI, Postgres, migrations, HTTPS proxy, and business tests.

Python API made with AI: moving FastAPI or Django from development to production

The development server is not the final destination

Demonstrations usually run with automatic reload, local credentials, and no error handling. That makes programming easier, but it is different from operating an API with users. FastAPI is typically served via ASGI and Django can use WSGI or ASGI; they are configured according to your real stack.

What to check before starting the API

Python and package versions, data access, migrations, environment variables, and access control. In Django, DEBUG=False and ALLOWED_HOSTS need attention. In FastAPI, check schema validation and error behavior. No application should connect to PostgreSQL with a superuser for normal operations. Avoid uploading a production .env to the repository.

Example: simultaneous reservations

AI-generated code checks available seats, saves a reservation, and sends a confirmation. Two people reserve the last spot at the same time and both see success. It is a concurrency error: it is not solved by adding memory to the VPS. Implement appropriate restrictions or transactions in the database, test two simultaneous requests, and verify that only one wins. The email is sent when the operation is confirmed.

Steps for a serious deployment

1. Identify the correct framework and command for production. 2. Pin dependencies, runtime, and private configurations. 3. Prepare PostgreSQL with versioned schema and backup. 4. Run ASGI/WSGI server with appropriate user and service. 5. Publish behind HTTPS proxy and configure logs and health checks. 6. Test roles, concurrency, migrations, and error recovery.

FastAPI or Django?

FastAPI is often convenient for typed APIs; Django provides integrated ORM, authentication, and admin. Both can serve professional applications. Choose based on current architecture and need, not on a prompt's preference.

Frequently asked questions

### Can I leave Django runserver in production? It is not the recommended server for production traffic. ### Is Uvicorn --reload for real users? Reload mode is for development; configure appropriate execution. ### What do I do with failed migrations? Test beforehand, back up the database, and design a recovery procedure. ### Does a VPS fix double booking? No: transactional logic and database constraints are needed.

Sources

FastAPI Deployment · Django Deployment Checklist.

The role of ArduMaker and the migration service

The ArduMaker infrastructure can host the project via VPS and compatible services, but it does not automatically convert a prototype into secure software. ardumaker.com/lovable is planned as a page to present a migration and deployment service for AI-generated code to production; it has not been created yet. We are not claiming that an automatic importer already exists. Each case requires code, dependency, data, domain, testing, and support review. The current platform can be consulted at ArduMaker.

Editorial review: October 2026. Check the requirements of each framework before making changes in production and keep restorable copies.

What the photograph shows

Availability status screen on a development machine. Illustrative photograph generated for this article about taking code created with AI to a production application; it is not an official capture of the tool nor does it demonstrate a specific implementation.