Product design question
Users of Perplexity Computer say they do not understand why the agent clicked, searched, or changed course, and they hesitate to let it continue. Design a core interaction model that makes the agent’s intent, current state, and next actions understandable and controllable without turning the product into a complex dashboard.
- Perplexity
- Product design
- Hard
Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.
Start a mock interview on this question · Mock interview from a job description
What this question tests
Tests core interaction design for agent legibility, making autonomous actions understandable and controllable without turning the product into a dashboard.
How to approach it
- Design a persistent, lightweight status indicator showing the agent's current action in plain language, for example 'searching flight prices' rather than a raw action log.
- Design an intent preview before higher-risk or hard-to-reverse actions, showing what the agent is about to do with a quick approve or redirect option.
- Design a compact history trail the user can expand on demand, so the full action log exists without being forced onto the main view.
- Design an always-available interrupt control, letting the user redirect or stop the agent mid-task without losing the session's progress.
- Keep the default view minimal, showing only current action and next likely step, reserving full detail for an opt-in expanded view.
- Test the design specifically against the stated user complaint, hesitation to let the agent continue, by measuring whether users let longer sequences run uninterrupted after the redesign.
What a strong answer includes
- Separates a minimal default view from an expandable detail view, solving legibility without defaulting to dashboard complexity.
- Adds an intent preview specifically before higher-risk actions, targeting the exact hesitation users reported.
- Defines a concrete success test, longer uninterrupted sequences post-redesign, tied directly to the original user complaint.
Common mistakes
- Designing a full always-on action log as the default view, recreating the dashboard problem.
- No interrupt control, leaving users unable to act on newfound understanding of agent intent.
Likely follow-up questions
- How would you handle an action so fast there is no time for a preview?
- How would you measure whether users actually trust the agent more after this redesign?
More product design questions
- How would you improve Perplexity's Comet browser for daily use?Perplexity · Product design · Medium
- Design an AI shopping assistant inside Comet.Perplexity · Product design · Medium
- Design a 'Deep Research' feature that produces trustworthy, citable reports.Perplexity · Product design · Hard
- How would you design Comet Assistant to safely act on tasks like booking flights and managing email?Perplexity · Product design · Hard
- Design the core interaction model for a Project where a human and multiple agents collaborate on ongoing work. How would you handle delegation, approvals, edit history, permissions, and recovery from bad agent actions so the experience feels powerful but dependable?Perplexity · Product design · Hard
- Perplexity notices that many users start a research task on web, ask several follow-ups, but do not return to that work on another platform later. How would you identify the single highest-leverage friction point in the begin → continue → return journey, and decide whether to invest first in context carryover, navigation, or collaboration?Perplexity · Product design · Hard
More questions from Perplexity
Learn the skill behind it
Chapters of the AI PM course that teach what this question tests.
- Chapter 4: Discovering and Sizing AI Opportunities
- Chapter 7: Designing AI UX and Oversight
- Chapter 14: Interviewing for AI PM Roles