Hackathon Finalist · 1st Place (Education Track)

PrepTalk

An AI-powered interview prep platform that analyzes résumés, generates role- and company-aware questions, and simulates realistic interview practice with structured feedback.

Role UX Design, Front-End Development
Team Hackathon team
Stack React, TypeScript, FastAPI, OpenAI

Personalized Questions

Generated interview questions using résumé content, role, and company context.

Mock Interview Flow

Created a guided prep and interview experience with realistic structure and feedback.

Accessible Prep Tool

Designed to reduce barriers for students who lack mentors or formal mock interview practice.

Overview

What the project does

PrepTalk is an interview preparation platform built to help users practice more effectively with personalized, structured support. Instead of relying on generic question banks or scattered resources, the product brings résumé analysis, company context, tailored interview questions, and mock interview feedback into one flow.

The goal was to create an experience that felt practical and immediately useful — especially for students and early-career candidates preparing for interviews without easy access to mentors, coaching, or realistic practice environments.

Problem

Why interview prep feels fragmented

Interview preparation is often spread across disconnected tools and resources. Candidates may search for likely questions, read company pages, revise résumés, and practice answers separately — but still lack feedback that feels personalized or realistic.

  • Practice is often generic instead of tailored to the user’s background
  • Company research and interview prep usually happen in separate places
  • Most students do not have access to realistic mock interviews or useful feedback
  • Technical and behavioral preparation are rarely integrated into one flow

The opportunity was not just generating questions with AI — it was designing a single product that made interview prep feel more personal, realistic, and actionable.

Solution

Designing an end-to-end prep experience

PrepTalk was designed as a guided system rather than a single-screen tool. The experience moves users from setup into preparation and then into interview simulation, helping them feel more prepared at each step.

  • Upload résumé and enter target role or company
  • Generate tailored interview questions
  • Review company-aware prep context
  • Run behavioral or technical mock interview flows
  • Receive structured feedback and improvement guidance

My Role

What I contributed

My work focused on UX design and front-end development. I helped shape the interview-prep flow, organize the experience into clearer stages, and build the interface so users could move from setup to practice in a way that felt intuitive and usable.

UX Design

Structured the flow into setup, prep, and interview stages to reduce confusion and support progression.

Front-End

Worked on the interface and interaction patterns that supported guided prep, question display, and user flow.

Product Thinking

Focused on how personalization, realism, and feedback could work together as one cohesive experience.

UX Decisions

How the experience was structured

One of the biggest design priorities was keeping the product from feeling overwhelming. Interview prep already carries stress, so the interface needed to feel clear, guided, and focused rather than overloaded with too many choices at once.

Step-Based Flow

Separated setup, prep, and interview stages to create a clearer mental model for users.

Personalization First

Used résumé, role, and company input as the basis for a more relevant prep experience.

Panelized Feedback

Organized outputs into readable sections so users could quickly understand next steps.

Behavioral + Technical Support

Designed the product to support multiple interview types instead of a one-size-fits-all flow.

Confidence Through Structure

Made the experience feel more approachable by reducing ambiguity and clarifying progression.

Usability Under Time Pressure

Prioritized clean layout and direct actions for a fast-paced hackathon build that still felt polished.

Demo

See it in action

Technical Implementation

How it was built

Frontend

React, TypeScript, and Vite for a responsive interface and structured client-side flow.

Backend

FastAPI used to support AI-powered interview generation and feedback workflows.

AI Layer

OpenAI-powered prompts were used to generate tailored interview questions and responses.

Technical Round

Monaco Editor supported coding-focused interview practice in the product experience.

A key part of development was making the AI outputs feel usable inside the interface, not just technically functional. That meant thinking carefully about prompt structure, result formatting, and how users would actually move through the prep process.

  • Prompt structure for consistent question generation
  • Front-end state and flow between prep stages
  • Readable display of tailored questions and feedback
  • Technical interview support through editor-based interaction

Outcome

Recognition and impact

PrepTalk was recognized as a finalist and received 1st Place in the Education Track. The project stood out because it paired a practical problem with a clear end-to-end solution: helping people prepare for interviews in a more equitable, personalized, and realistic way.

  • Finalist recognition at the hackathon
  • 1st Place in the Education Track
  • Combined UX, AI, and front-end execution in one product
  • Addressed a real accessibility and opportunity gap in interview preparation

Takeaways

What this project taught me

PrepTalk reinforced how valuable it is to design AI-powered products around usability and trust. The strongest part of the project was not just the AI generation itself, but how that intelligence was shaped into a product flow that felt practical, supportive, and easy to use.

  • Designing AI experiences that feel clear and useful
  • Structuring guided flows for higher-stress tasks
  • Building front-end systems around dynamic generated content
  • Balancing speed of implementation with usability in a hackathon setting