Tool docs that helped me: model evaluation↗ openai.com
Sharing a resource related to prompt engineering 003.
- Why it matters: overview + practical tips.
- What I’m curious about: do you agree with the recommendations?
41
Sharing a resource related to prompt engineering 003.
Agree. Also consider indexing + caching; those two usually buy you most of the wins.
If you share constraints (latency, budget, scale), it’s easier to recommend.
One more thought: validate assumptions with a small A/B test if possible.