Inspiration We noticed that while people spend hours perfecting resumes and applying for hundreds of jobs, they rarely prepare enough for interviews. Many don’t have access to mentors or professionals who can give real feedback. Practicing with friends or roommates often feels awkward, unhelpful, or even discouraging. We wanted to create a safe, intelligent, and supportive space where job seekers can practice interviews confidently and get both AI-driven insights and human expert feedback.
You can try it here : https://61bb7c5c6f65.ngrok-free.app/
What it does InterviewAI is a dual-approach platform that combines AI-powered mock interviews with a human expert network: Users practice real-time mock interviews with AI using speech-to-text and natural language processing. They get instant performance analytics on technical skills, communication, and problem-solving. They can then connect with industry professionals for one-on-one mock sessions, personalized feedback, and even networking opportunities. A complete dashboard helps track progress over time, providing tailored recommendations to improve.
How we built it
Frontend: React + TypeScript, TailwindCSS, shadcn/ui for modern, responsive design. Backend & Services: Firebase (Firestore, Auth, Storage, Hosting) for secure and scalable infrastructure.
AI & Voice Tech: OpenAI GPT-4o-mini for intelligent question generation and analysis OpenAI Whisper API for accurate speech-to-text 11Labs voice AI for natural speech interaction WebRTC for real-time voice recording and streaming Dev Tools: Vite for fast builds, ESLint/Prettier for clean code, React Query for state/data handling, and Recharts for analytics visualization.
Challenges we ran into
Integrating real-time speech analysis with Whisper and ensuring low latency. Designing an analytics framework that feels meaningful rather than generic. Handling secure authentication and data privacy, since we store personal interview data. Balancing between AI-driven practice and human professional engagement without making one feel less valuable.
Accomplishments that we're proud of
Building a working AI interview engine with voice-based interactions. Creating a community platform for human experts, complete with scheduling and payments. Designing a clean, intuitive user experience that makes interview practice engaging, not intimidating. Achieving an end-to-end functional prototype where users can log in, practice with AI, book experts, and view analytics.
What we learned
How to combine speech AI + NLP + web technologies into a seamless product. The importance of data visualization in motivating users—clear progress tracking makes users want to improve. That building trust with users requires strict security practices and transparency in how we handle interview data. How to think like both job seekers and recruiters, bridging the gap between practice and real-world interviews.
What's next for InterviewAI
Mobile App Launch: Expanding to iOS/Android for on-the-go practice. Role-Specific Playbooks: Tailored interview flows for software engineers, data scientists, product managers, etc. Gamification: Streaks, badges, and leaderboards to keep practice fun. AI Feedback Fine-Tuning: Making analysis more contextual (e.g., for system design vs. behavioral interviews). Corporate Partnerships: Offering the platform to universities, bootcamps, and companies as part of their career services.
Log in or sign up for Devpost to join the conversation.