You are asked to evaluate a new ML feature before launch. The team wants to know whether the feature improves the main metric enough to ship, and whether any guardrail metrics move in the wrong direction.
How would you evaluate the impact of a new ML feature before launch?
You are asked to evaluate a new ML feature before launch. The team wants to know whether the feature improves the main metric enough to ship, and whether any guardrail metrics move in the wrong direction.
How would you evaluate the impact of a new ML feature before launch?
