PrepLens

Elevate your interview preparation through an AI interviewer

Comment

Inspiration

The job interview process can be daunting, with candidates often struggling to bridge the gap between theoretical preparation and real-world application. We were inspired to create a solution that not only simulates realistic interview scenarios but also provides personalized, actionable feedback to boost confidence and readiness. By combining AI, speech recognition, and facial analysis, we aimed to build an all-in-one tool to revolutionize interview preparation.

What it does

PrepLens is an AI-powered interview preparation platform that:

  • Simulates realistic mock interviews with AI-generated questions tailored to job descriptions.
  • Converts user responses from speech to text for analysis.
  • Evaluates answers and provides feedback with ratings to improve performance.
  • Analyzes resumes, offering ATS (Applicant Tracking System) scores and highlighting missing keywords.
  • Monitors user facial expressions and posture during interviews, giving insights on non-verbal communication.
  • Alerts users if they move out of frame during the interview.

How we built it

  • Frontend: Developed with Next.js, Typescript, and shadcn-ui for a seamless and responsive user experience.
  • AI Integration: Used Gemini API for question generation, answer feedback, and ATS scoring.
  • Speech Recognition: Integrated React Speech-to-Text for converting spoken answers into text.
  • Facial Analysis: Incorporated face-api.js to track user expressions, posture, and focus.
  • Backend: Implemented Prisma with Supabase for database management and real-time updates.

Challenges we ran into

  • Calibrating speech-to-text accuracy for different accents and speaking speeds.
  • Implementing real-time facial analysis while maintaining performance and user experience.
  • Parsing AI-generated feedback into actionable, structured JSON responses.
  • Ensuring the ATS scoring algorithm aligned with industry standards.

Accomplishments that we're proud of

  • Successfully integrated multiple technologies to provide a holistic interview preparation experience.
  • Created a user-friendly interface that feels intuitive and engaging for users.
  • Achieved accurate and meaningful feedback from the AI model for both answers and resumes.
  • Developed a robust system for facial analysis that tracks posture and alerts users dynamically.

What we learned

  • The importance of real-time processing for user interactions, especially in mock interviews.
  • Techniques for optimizing AI API calls to minimize latency without compromising accuracy.
  • Deeper insights into facial detection technologies and how non-verbal cues impact interviews.
  • How to design a modular, scalable system that integrates diverse technologies seamlessly.

What's next for PrepLens

  • Enhanced Insights: Incorporate more advanced analytics like tone detection and emotional sentiment analysis during answers.
  • Diversity in Questions: Expand the question bank to include industry-specific and behavioral interview questions.
  • Mobile App Support: Launch a mobile app version for greater accessibility.
  • Gamified Learning: Add interactive quizzes and challenges to make preparation more engaging.
  • AI Mentorship: Introduce an AI mentor that guides users through their preparation journey, offering personalized tips and strategies.

Built With

Share this project:

Updates