Run a pre-mortem before launch
Use a pre-mortem to identify and reduce likely AI workflow launch failures.
A scheduling agent is ready for executive assistants. The demo works, but the team has not tested social and calendar edge cases. Assume the launch failed, then work backward to prevent the causes. The novice move is to ask, does anyone see risks? People stay polite, optimistic, and abstract. Before Risk review: The team says adoption risk is medium and calendar permissions should be fine. After Pre-mortem: Assistants abandoned the agent because it moved VIP meetings, ignored private holds, and had no visible undo path. Set the scene It is 30 days after launch. The scheduling agent has failed and…
Sign up free — one personalized lesson every day, matched to your role and goals.
Already have an account? Sign in