Run an after-action review after a bad AI run
Conduct a short after-action review that improves a prompt workflow without blame.
A market brief went to leadership with a stale source quote. The team is frustrated and wants a fast fix before next week's brief. AAR: expected -> actual -> why different -> sustain or improve The common trap is ending the review with "review more carefully" and no change to the prompt, source labels, or checklist. Before Incident response: "The reviewer should have caught the quote. Add a reminder to be accurate." After AAR response: expected only current packet quotes; actual output used a stale deck; cause was unlabeled source age; improvement is date-and-authority labels plus a refusal rule for…
Sign up free — one personalized lesson every day, matched to your role and goals.
Already have an account? Sign in