Behavioural: "Tell me about a time you pushed back on a customer." How should you structure your answer, and what does a strong answer sound like?

Situation: "At a mid-size insurer, the operations head wanted our claims assistant to auto-approve all claims under ₹50,000 from week one of the pilot to hit a cost-reduction target."

Task: "I owned the deployment, and I believed auto-approval without validation could cause wrong payouts and regulatory issues."

Action: "I ran our eval set on their last three months of claims and found 7% disagreement with adjusters on that segment, mostly on documents with missing hospital stamps. I showed him what that meant: about 280 potentially wrong approvals a month. Instead of just saying no, I proposed a phased plan: AI recommendations with one-click adjuster approval for four weeks, then auto-approval only for the claim types where agreement exceeded 98%. I looped in their compliance lead early and documented the decision."

Result: "He agreed. Adjusters' review time dropped 40% in the first month, and by week eight we enabled auto-approval for two claim types covering 35% of volume with no reversals in the next quarter. The operations head later asked us to use the same phased approach for another line of business."

Learning: "Pushback lands better with their own data and a path to their goal, not a flat no."