Interview Question for Web Developer at Mozilla:

This is fairly open-ended, and it depends a lot on the type of application, the server achitecture, the expected load, etc.

Some possibilities:
- Find slow requests by examining the web server or app server logs.
- Profile the requests using the profile/cProfile python modules.
- Optimize any queries the requests are performing or the database schema the queries are executed against. This depends a lot on the database being used.
- If possible, move expensive create/update/delete tasks into a background job queue and out of the request/response cycle.
- Add more app servers or database servers.
- Apply more caching, particularly of expensive retrieval tasks.