Pick the prototype when the model is not ready

The model is six weeks away; the user-behavior risk is due this week. What should Lena prototype to learn whether merchandisers can use the AI recommendation workflow? Run a clickable workflow with human-prepared recommendations behind the scenes. Right. The workflow and recommendation examples are realistic enough to reveal trust, edits, and objections without waiting for model implementation. Wait until the ML model works so every test recommendation is authentic. Weaker. That delays product learning even though the current risk is interaction behavior, not model accuracy. Create a polished video demo for executives and infer user trust from their reaction. Weaker.…

Sign up free: one personalized lesson every day, matched to your role and goals.

Already have an account? Sign in