IntPrepAI
Customer Journey Design
Designing the end-to-end user experience for an AI-powered job interview preparation platform, from onboarding and persona setup through to mock interviews, CV gap analysis, and readiness tracking.
Role
Product Designer
Timeline
Dec 2025 – Present
Location
Hong Kong
Tools
Figma · Google Stitch · Miro
The Challenge
Job interviews are one of the highest-stakes moments in anyone's career, yet most candidates prepare alone, without feedback, and without knowing what to expect. IntPrepAI needed to make AI-powered coaching feel accessible, not intimidating.
The UX challenge: design an experience that delivers genuine value within the free tier to earn the upgrade, without overwhelming a first-time user before they've even answered a question.
The Approach
I mapped four distinct user journeys, each representing a different mode of the app, to show how a user progresses from first launch to interview-ready confidence across onboarding, mock interviews, CV analysis, and cheat sheet generation.
Each journey was visualised using the Obsidian Navigator design system in Google Stitch: dark-mode, editorially structured, and built to communicate the app's value to both candidates and stakeholders.
The Four Customer Journeys
Each journey maps a distinct mode of the app, from the first-launch setup flow through to the Free-to-Premium conversion point.
Onboarding & Job Setup
The welcome screen presents all four core capabilities upfront before the user commits: job-specific questions, CV gap analysis, scored practice, and a readiness tracker. A single "Get Started" CTA routes every user through the job spec upload, setting personalised context for the entire app.
Four pillars surfaced immediately, Job-Specific Questions, CV Gap Analysis, Practice & Get Scored, and Know When You're Ready shown before sign-up
Single entry point, one "Get Started" button routes all users through the same job spec upload, ensuring the app is personalised from the first screen
Role-specific framing, every feature description references the job spec, setting the expectation that all content is tailored to the user's target role
AI Mock Interview Session
The live mock interview puts the candidate face-to-face with a named AI persona — here, an Engineering Manager from AGI Data Services. The AI poses a real technical question drawn from the job spec and evaluates the response across Tech dimensions including Technical Accuracy and Problem-Solving Approach. The candidate answers by voice, with options to switch to text, request a hint, or skip.
Named AI persona, the interviewer takes a specific role from the target company, reinforcing job-specific context throughout the session
Voice-first with fallback, recording is the primary input method, with Type Instead available for candidates who prefer text
In-session support, Get Hint and Skip Question give candidates a lifeline without breaking the session flow
CV Gap Analysis
CV Optimization scores the uploaded CV against the job description with a single match percentage — 75% here — then splits the analysis into two clear sections. "What's Working" confirms the strengths the interviewer will value, while "Areas to Improve" flags specific gaps with concrete suggestions, from missing LLM experience to enterprise database familiarity.
Match score upfront, a single percentage gives immediate context before the user reads any detail
Strengths confirmed, not just gaps, five "What's Working" items build confidence before surfacing what's missing
Actionable gaps, each "Areas to Improve" item is specific: "Add specific experience with Large Language Models and multimodal systems", not generic advice
Cheat Sheet & Progress Tracking
The Cheat Sheet distils everything into a single, role-specific reference document. Company Highlights surface what the interviewers care about — LLM training integrity, the Amazon Nova model line, and the AGI team's evaluation mandate. "Your Strengths to Emphasise" then maps the candidate's own CV directly to those priorities, so every talking point is grounded in something they've actually done.
Role header at the top, company, team, and interviewer shown immediately so the candidate knows exactly who they're preparing for
Company intelligence, three highlights about Amazon's AGI team pulled from public context about the Nova models and LLM evaluation mission
CV-matched talking points, strengths are phrased in the company's own language, bridging the candidate's experience directly to the team's stated priorities
Design Outcomes
Onboarding, mock interview, CV analysis, and cheat sheet, each mapping a different mode of the app.
AI Personas, Mock Interviews, Score Breakdowns, CV Gap Analysis, Cheat Sheets, and Progress Tracking all visualised end-to-end.
Conversion point placed at peak value, after the user has seen measurable progress and exhausted free tier features.
Interested in this project?
Let's talk about how I can bring this kind of thinking to your product.