Product design question

Design a feature to automatically validate the identity of someone that is applying for a loan. How much does this feature cost?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Product design combined with cost reasoning for a fraud-sensitive identity verification feature in lending.

Clarify the risk being solved, identity theft, synthetic identity fraud, or income misrepresentation in loan applications.
Design the verification flow, government ID upload with liveness check via selfie, cross-referenced against a credit bureau or identity database.
Balance friction against security, keeping the flow to a few minutes since excessive steps cause drop-off in loan applications.
Estimate cost components, per-verification API fees for ID and liveness checks, typically charged per attempt by third party vendors.
Estimate volume-based total cost, cost per verification multiplied by expected monthly loan applications, as an illustrative calculation.
Define success metrics, fraud rate reduction and verification completion rate, balanced against the added cost per applicant.

Breaks the feature into concrete steps, ID upload, liveness check, and database cross-reference, rather than a vague 'verify identity' claim.
Shows a real cost calculation, per-verification fee times expected volume, clearly marked as an illustrative assumption.
Balances fraud reduction against application drop-off, since too much friction in verification directly costs loan volume.
Sets a dual metric, fraud rate and completion rate, showing the security versus friction tradeoff explicitly.

Proposing verification with no cost estimate despite the question asking for one.
Ignoring the tradeoff between security friction and application completion rate.
Not addressing what happens when verification fails, e.g. a manual review fallback.

Chapters of the AI PM course that teach what this question tests.