GPT Assistants API with NGINX as reverse proxy

Hi, everyone!

I’ve developed a bot and an API for Assistants to be used in a service at my company.

I’m looking to control traffic by implementing caching to reduce overall costs. Additionally, I’m interested in leveraging features like rate limiting offered by a reverse proxy.

Another consideration is implementing different versions for various users through a canary deployment. My current plan involves using Nginx to facilitate this layer.

Has anyone experimented with similar setups and would like to share their experiences?