EchoPrep

EchoPrep is a mobile-first voice-based mock interview app built with React Native. Users can practice articulation, refine answers, and receive AI real-time feedback.

Comment

Inspiration

I’ve always struggled with acing interviews, and most available solutions are either expensive or lack personalisation. I wanted a tool that could simulate real interviews, give honest feedback, and adapt to my needs. This hackathon gave me the perfect opportunity, not just to solve a personal pain point, but also to finally build a mobile app from scratch and participate in my first hackathon.

What it does

EchoPrep is an AI-powered mobile app that helps users prepare for interviews through personalised, voice-based mock sessions. By analysing your resume and job description, the app generates tailored questions and reads them aloud using text-to-speech (TTS). You respond verbally, and the app transcribes your answers using speech-to-text (STT), then evaluates them to provide instant feedback. Each follow-up question considers your earlier responses for a more dynamic and realistic session. After the interview, you receive a detailed summary and can export the feedback in markdown format for future review.

Key features:

🎯 Personalised Interview Sessions Upload your resume and job description (or let the AI generate one). EchoPrep then tailors questions based on your background and target role. You can fine-tune the session by adjusting the mix of technical and behavioural questions.

🎙️ Voice-Powered Practice Experience natural, immersive mock interviews. Questions are read aloud, and your spoken responses are transcribed in real-time.

🤖 Intelligent Feedback Engine Get real-time feedback on your answers, assessing clarity, content, confidence, and delivery. At the end, receive a final articulation score with a breakdown of strengths and areas for improvement.

📄 Resume & JD Management Easily upload and parse resumes (PDF/text), and save multiple job descriptions. The AI ensures questions remain relevant to your experience and goals.

📝 Smart Notes & Markdown Editor Save AI feedback and personal notes using the built-in markdown editor. Track your progress and prepare more effectively for future interviews.

How we built it

  • Before the hackathon, I brainstormed ideas and scoped the MVP with help from ChatGPT.
  • Built the app using React Native (via Expo) and Supabase for backend services (auth, storage, edge functions).
  • Developed core features iteratively, starting with resume parsing and question generation.
  • The interview session module was the most complex, combining TTS, STT, and dynamic question logic.
  • Implemented supporting features like notes and session history.
  • Implemented secure token-based usage using Supabase edge functions and RevenueCat for in-app purchases.

Challenges we ran into

  • First time using React Native, Supabase, Expo, and edge functions, all within a tight 30-day deadline.
  • RevenueCat does not work in Expo Go, which meant testing had to be done in standalone builds only.
  • Bolt.new’s token limits made debugging and iteration costly once the app grew, each test run used ~200,000 tokens.
  • Managing build errors and React Native quirks, especially on Android emulator, while preparing a clean demo.

Accomplishments that we're proud of

  • Built and launched a fully functional mobile app from scratch in 30 days with no prior mobile dev experience.
  • Created a voice-based mock interview experience that adapts to real user input and provides useful, personalised feedback.
  • Gained hands-on experience with Supabase edge functions, speech processing APIs, and mobile in-app purchases.

What we learned

  • Summarising and chunking context is essential to keep AI interactions within token limits.
  • Learned the entire Google Play Store publishing pipeline, including key management, testers setup, and AAB signing.
  • React Native’s limitations and workarounds, especially for native modules and build tools.

What's next for EchoPrep

  • Complete deployment to Google Play Store (currently testing phase) and start Apple App Store submission.
  • Polish UX/UI and improve feedback quality with sentiment and filler word analysis.
  • Expand content options (e.g., role-specific technical questions) and allow users to customize voice personas.
  • Market the app to job seekers, fresh grads, and bootcampers looking for affordable and accessible interview practice.

Built With

Share this project:

Updates