startup page
join a room
join a meeting
create a meeting
login to your account
In a room with three people
interactive and synced IDE
Use AI to enhance low resolution stream
Live transcription and translation
AI Correcting answers

ViewInter

Real-time video interviews with AI-powered transcription and translation

Inspiration

Technical interviews shouldn't be limited by language barriers or lack of documentation. ViewInter bridges the gap between interviewers and candidates worldwide with seamless real-time communication tools.

What it does

ViewInter is a full-featured video conferencing platform purpose-built for technical interviews:

Peer-to-peer video calls with WebRTC for crystal-clear, low-latency communication
Real-time speech-to-text powered by Mistral's Voxtral API with automatic language detection
Live subtitles rendered directly on participant video feeds
Shared code editor instantly synced across all participants for live coding challenges
Smart Voice Activity Detection that intelligently transcribes only when you're speaking
Automatic transcript generation saved for comprehensive post-interview review
Full user authentication with a polished meeting management dashboard

How we built it

Layer	Technology
Frontend	Angular 19, Material Design
Backend	Node.js, Express, Socket.IO
Video	WebRTC, PeerJS
AI/ML	Mistral Voxtral API, ONNX Runtime (Silero VAD)
Database	MongoDB Atlas
Hosting	DigitalOcean App Platform

Technical Challenges

Browser-based VAD: We integrated the Silero VAD model running entirely client-side via ONNX Runtime WebAssembly, enabling instant speech detection with zero server round-trips
Real-time synchronization: Coordinated Socket.IO events handling video streams, transcriptions, code changes, and user presence all simultaneously
Latency optimization: Carefully debounced transcription requests and optimized audio chunking to deliver responsive, natural-feeling subtitles
Cross-origin architecture: Engineered separate static frontend and WebSocket-enabled backend deployments with robust CORS handling