Overview
Project: AI Interview Practice Tool
Role: Lead UX Designer
Platform: Monster Mobile App (iOS & Android)
Team: Product, Mobile Engineering
Tools: Figma, OpenAI, User Testing.com
The Opportunity
Practicing interviews is inherently awkward and hard to do alone. AI could help, but only if users trusted it and understood how to interact with it.
Our Goals
• Help users prepare for interviews based on actual job descriptions
• Improve user confidence and performance
• Drive re-engagement with the Monster app post-application
The Feature
We designed an AI-powered mock interview tool with the following flow:
• User selects a job they applied to (or is interested in)
• AI generates personalized interview questions based on the job title & skills
• User records their audio responses in-app
• AI evaluates their delivery, pacing, tone, and content
• Users receive a score + specific feedback + improvement tips
Why AI Could Fail
• Early concepts risked feeling either too robotic or too open-ended, making users unsure how to engage or whether the feedback was meaningful.
• Encouraging users to try something high-friction (audio recording)
• Creating a non-judgmental, coaching-like tone in AI feedback
• Balancing detail and simplicity in the results screen
• Designing for privacy and control over recordings
Process Highlights
• Early concept testing validated user interest in “interview practice” but flagged performance anxiety — we emphasized that this was private, not shared
• I chose to structure the experience as guided sessions rather than open-ended interaction to reduce ambiguity and give users a clear starting point.
• I intentionally avoided adding advanced customization early on to keep the experience simple and approachable.
Added features like:
• Re-record option before submission
• Toggle to pick a “difficulty level”
• Visual cue during AI “scoring” phase (animated waveform + loading)
• Worked closely with the AI team to ensure feedback categories were human-readable: e.g., “You used a lot of filler words — try to pause instead.”
Results (Initial Rollout Metrics)
• 18% of job applicants tried the tool within 72 hrs of applying
• 4.5 / 5 average rating for the usefulness of feedback
• 63% completed at least one full mock interview
• Re-engagement increased by 21% for those who used the tool
Reflection
This feature pushed me to think about AI not as a gimmick, but as a coach. We learned that users appreciated feedback, but preferred tone coaching and pacing tips over keyword/skill analysis. In future iterations, I’d explore:
• Adding a “review past interviews” feature
• Letting users choose focus areas (e.g., STAR method, confidence, articulation)
• Real-time coaching (e.g., “Try rephrasing that” mid-interview)
