Projects using WebRTCProjects using WebRTC
Projects
People
Cover image for Converc — Turn Website Visitors
Converc — Turn Website Visitors Into Live Conversations URL: https://converc.com One-liner Built a real-time browser-based calling platform from scratch — MVP live in 5 days. Project Overview Converc solves a problem that costs B2B companies revenue every day: high-intent visitors land on your site, can't get a human immediately, and leave. It embeds a call widget that connects visitors to sales reps in real time — no booking links, no forms, no friction. I architected and built the full product: a Next.js frontend, Supabase backend with Row Level Security, and WebRTC-powered peer-to-peer audio calling that works entirely in the browser with no native app required. The stack was chosen for performance, low latency, and the ability to ship fast without sacrificing production quality. What I Built WebRTC call engine with real-time signalling via Supabase Realtime Embeddable call widget (drop-in script for any website) Agent dashboard with live call status, queue management, and session history Webhook integration layer for CRM and Slack notifications on call events OAuth-based authentication with Google and dev/prod environment separation Supabase RLS policies enforcing strict data isolation between workspaces How I Shipped It The core MVP — widget, signalling, agent dashboard, and working calls — was shipped in 5 days using Cursor with Claude as the AI pair programmer. Post-MVP work covered Google Safe Browsing clearance, Slack App submission, analytics scoping, and hardening the auth and security model for production. Tech Stack Next.js · Supabase · PostgreSQL · WebRTC · Vercel · Tailwind CSS
0
41
Cover image for AI-Powered Algorithmic Trading Platform —
AI-Powered Algorithmic Trading Platform — AI Systemics Built core technology infrastructure for AI Systemics — an AI company developing intelligent solutions that bring institutional-grade capabilities to everyday users. Primary focus on their flagship product, Quanntick™, an AI-powered algorithmic trading platform that brings institutional-level trading to individual investors. Developed the machine learning pipeline powering real-time market analysis and signal generation — processing live CME Group licensed market data, running predictive models, and executing automated trading strategies on futures markets with precision and speed. Built the performance tracking dashboard giving users transparent visibility into trade execution, strategy performance, and portfolio metrics in real-time.
1
3
309
Cover image for Mitsi MVP is LIVE! 🎉 I’m super excited to finally share thi...
Mitsi MVP is LIVE! 🎉 I’m super excited to finally share this! Over the past few months, I have been building Mitsi — an open-source video conferencing platform designed to be fast, modern, and user-friendly. It’s been an amazing journey exploring different technologies and crafting a real-time communication experience from the ground up. Here are some of the core technologies powering Mitsi: ● React – powering the Mitsi web client ● Mediasoup Client – the WebRTC engine on the frontend ● Mediasoup SFU – handling media routing with low-latency performance ● SocketIO – Mitsi’s custom signaling protocol between client and signaling service ● gRPC – enabling communication between the Mitsi media service and signaling service ● Nodejs – backend services ● Redis – service state sharing, caching, and Pub/Sub Link is in the comment
1
0
50
Cover image for GYMYG — Virtual Fitness Platform California,
GYMYG — Virtual Fitness Platform California, USA · 12 Months · 7-Person Team The Problem The home fitness market exploded but the tech never caught up. Existing solutions were either pre-recorded content with no real interaction, or basic video calls with no fitness intelligence built in. GYMYG needed something that didn't exist yet — a platform where certified trainers and clients could train together in real time, with the intelligence to match them, program for them, and keep them accountable long after the first session. What We Built A complete virtual fitness ecosystem — two separate applications for trainers and clients, coordinated through a single backend, with an admin layer running everything underneath. Sessions run over WebRTC with adaptive bitrate streaming, so video quality adjusts automatically based on connection strength without dropping the call. TURN/STUN resilience means sessions stay live even on difficult networks. In 50,000+ sessions delivered, the failure rate never crossed 1%. The AI layer does the work that makes GYMYG more than just a video call. Programming is calibrated to each user's goals and adherence patterns — if someone keeps skipping leg day, the system knows. Biometric and session telemetry track safe progression and flag recovery needs before they become injuries. Trainer-client matching uses availability, specialty, and historical satisfaction signals — not just whoever's online. Community features — live group classes, asynchronous content, and moderated chat — were built in from day one because retention data is clear: people who train alone quit. People who train in community don't. Technical Architecture Real-time sessions: WebRTC with adaptive bitrate streaming + TURN/STUN resilience Backend: Node.js services layer handling real-time state, scheduling, and payments Database: MongoDB with rigorously designed schemas and TTL-based event retention Infrastructure: AWS with autoscaling, managed queues, and full-stack observability — metrics, logs, traces Security: Tokenized access, scoped permissions, encrypted media signaling throughout Stack: React Native · Node.js · WebRTC · MongoDB · AWS Results 50,000+ live sessions delivered 95% customer satisfaction across rolling quarters Sub-1% session failure rate Improved 90-day retention driven by social features and AI-personalised programming First live fitness app to offer real-time trainer-client interaction at this scale Successfully launched into the US fitness market in 12 months What This Proves Real-time platforms at scale aren't just a backend problem. They're an architecture problem. The decisions made in month one — how sessions are handled, how data is structured, how the AI layer sits inside the product — are what determine whether you hit 50,000 sessions or fall apart at 500. GYMYG was built to handle the former from day one.
0
38