Step 4 of 9 · planned
Fix the prompt, prove it
Change the prompt with evidence instead of vibes, and find out what the fix broke.
What you'll build
A second prompt version, a second experiment on the same dataset, and a row-by-row comparison.
Braintrust features
What you'll be able to show
“v1 against v2, row by row — including what got worse.”
🐾 Not written yet. Each step is written as the course is actually worked through, so it describes real scores, real regressions and real false alarms rather than a script guessed in advance.