Proving depth of experience with LLM-powered and other AI products in a PM interview

In a product manager interview at Accenture, you are asked whether you have hands-on experience with AI or generative AI products. Answer as you would in the interview. ```hint Go past yes or no It sounds like a screening question, but the interviewer is looking for depth. Pick one product, and be ready to explain what made the AI part different from building ordinary software. ``` ### Clarifying Questions - Does the interviewer mean products where I owned the AI capability, or also products where I integrated a third-party model or used AI tools in my own work? - Should I go deep on one product or give a short overview of several? - Because Accenture delivers much of its AI work for clients, should the example show work for an external client, or is in-house product work equally relevant? ### What a Strong Answer Covers - A direct answer followed by one specific example with the candidate's own role, the user problem and why AI was the right approach - Decisions particular to AI products: model choice, data, evaluation, cost and latency, and handling wrong or unsafe output - Measured outcomes for users and the business, plus what did not work - Honest scoping of experience and, where it is thin, credible adjacent evidence and a plan to close the gap - Relevance to delivering AI work for clients, such as educating stakeholders, responsible AI and moving from pilot to production ### Follow-up Questions - How did you know the model was good enough to launch, and who signed off on that bar? - Tell me about a time the AI feature produced a wrong or harmful output in front of users. What did you change? - A client wants a generative AI pilot but has no clear use case. How would you help them choose one?