Live
IntellMeet
A live AI-powered meeting and collaboration platform for scheduling, browser meetings, real-time chat, transcripts, summaries, action items, tasks, and analytics.
Problem
Remote meetings often scatter scheduling, live discussion, transcripts, summaries, tasks, and analytics across too many tools.
Scope
Students, small teams, and collaborators running structured online meetings.
My contribution
Built the meeting workspace across authentication, scheduling, room interactions, real-time chat, captions, recordings, summaries, action items, task management, analytics, and deployment.
- Implemented meeting scheduling, protected room flows, participant surfaces, chat, captions, recordings, and host controls.
- Connected live collaboration with Socket.IO, WebRTC-oriented room interactions, API state, and authenticated user sessions.
- Built post-meeting productivity features including AI-style summaries, action items, analytics, notifications, exports, and task boards.
Technical decision
Separated persistent meeting records from real-time room state so the collaboration layer could stay responsive while meeting artifacts remained queryable.
Architecture and data flow
Built the meeting workspace across authentication, scheduling, room interactions, real-time chat, captions, recordings, summaries, action items, task management, analytics, and deployment.
React and Vite frontend, Node.js and Express API, MongoDB persistence, Socket.IO event channels, JWT sessions, WebRTC-oriented meeting flows, Vercel frontend hosting, and Render backend hosting.
Authenticated users create and join meetings through APIs; live room events move through Socket.IO; meeting artifacts are stored for summaries, actions, analytics, and follow-up tasks.
Uses JWT-protected workflows and role-aware meeting controls; public claims are limited to the implemented app and verified live deployment.
Technical challenges
- Keeping real-time interaction responsive while saving meeting artifacts for later review.
- Presenting summaries, action items, analytics, and tasks without overwhelming the meeting workspace.
- Coordinating separate frontend and backend deployments with stable public routes.
Outcome
Live frontend and backend endpoints are available, with the project presented as a deployed portfolio product rather than a guaranteed production-scale service.
Limitations
Long-term uptime, enterprise-scale load, and independent security audits are not claimed.
Future improvements
Add deeper collaboration testing, improve meeting reliability under load, and expand post-meeting intelligence workflows.