- IAM for LLMs
- Not every user can ask any question
- Prompt engineering evaluation package
- In figuring out which is a good prompt, you may need to employ classification evaluation techniques, so that’s something that can be automated or scripted potentially