“You can get a surprisingly long way with an LLM and a quick prompt, but getting good results, consistently, is where the real work is.”
We wrote about building AI flag cleanup. From finding the right model, to ending the whitespace problem.
2 months ago