At Axflow, we are building open source tooling to help developers integrate large language models into their products. The majority of this tooling thus far has been written in TypeScript, though we've been exploring replacing part of our stack with a Phoenix application. We spend quite a bit of time working with OpenAI and given LLMs are all the rage, I wanted to write about how we're integrating with it in our Phoenix application.

This post is the first in a series of posts about integrating with OpenAI's chat completions endpoint.

continue reading on benreinhart.com

⚠️ This post links to an external website. ⚠️