Portfolio
HomeAboutProjectsContact

Mixora – AI-Enhanced Video Conferencing Platform

A next-gen video conferencing platform featuring automatic transcription, AI-powered summaries, and an industry-first threading system for maintaining context across multiple meetings. Built on Next.js 16, TypeScript, Prisma, and React 19 for a seamless and intuitive smart meeting experience.

Dec 2025
1 Repository

Technology Stack

Next.js 16
React 19
TypeScript
Tailwind CSS
Radix UI
Prisma ORM
PostgreSQL
AssemblyAI
Groq Llama 3.3
Stream.io SDK
NextAuth.js
Resend

Key Features

Smart video conferencing with automatic AI-powered summaries
Threaded meetings for context continuity
Secure, multi-role authentication (JWT/OAuth/Google)
Instant recording/playback and download
Visual thread tree for project evolution and decisions
Dashboard with meetings, recordings, summaries
Free, automatic transcription
Fully mobile-responsive modern UI

Tags

Next.js
React 19
TypeScript
Prisma
PostgreSQL
AssemblyAI
Groq Llama 3.3
Stream.io
Radix UI
Tailwind CSS
Resend
NextAuth.js

Project Details

Mixora revolutionizes professional meetings with AI-powered context awareness — going beyond simple video calls by connecting related meetings in "threads", transcribing conversations automatically, and generating smart, context-rich summaries. Never lose track of project evolution, decisions, or context across meetings again.


System Architecture

Full-Stack, Real-Time, AI-Enhanced

  • Frontend: Next.js 16 (App Router), React 19, TypeScript, Tailwind CSS, Radix UI
  • Backend: Next.js API Routes, Prisma ORM, PostgreSQL, JWT-based Auth
  • Real-Time Video: Stream.io SDK for secure, scalable video calls
  • AI & Transcription: Groq (Llama 3.3) for context-aware summaries; AssemblyAI for automatic transcription
  • Email & Scheduling: Resend for robust notification flows
  • Dashboard: Displays all meetings, recordings, summaries & thread structures

Intelligent Meeting Flow

User Journey

  1. Create or Schedule Meeting: Instantly or at a future time; invite by email/code
  2. Join Video Call: Browser-based, no downloads needed; screen share, record
  3. Recording & Transcription: Meetings are recorded; speech is auto-transcribed (multi-language)
  4. AI-Powered Summaries: Summaries highlight decisions, key points, action items, and include context from "threaded" meetings
  5. Threading System: Link follow-up meetings, all context is automatically preserved and mapped visually in a tree
  6. Dashboard: Browse all past meetings, threads, summaries, and manage past/future events

Signature Features

AI & Automation

  • Auto-Transcription (AssemblyAI): Every meeting is converted to text, accuracy 95%+
  • Context-Aware Summaries (Groq Llama 3.3): Summaries aren't isolated; they track decision changes, repeated topics, and provide a living timeline for each project or team
  • Follow-up Linkage: Meetings are threaded (parent/child) - each follow-up includes full context automatically

Security & Scalability

  • NextAuth.js with JWT/OAuth: Multiple auth options, secure password hashing (bcrypt), Google OAuth
  • Role-Based Access: Host, participant, and co-host privileges
  • Data Protected: End-to-end encrypted video, secure API endpoints

Best-in-Class User Experience

  • Modern UI: Built with Tailwind CSS, Radix UI, fully mobile-responsive
  • Visual Threads: Meetings displayed in tree view for project continuity
  • One-click Recording & Playback: Instant review of past meetings

Technical Implementation

Frontend

  • Next.js 16 + React 19: Fast, modular SPA with server rendering
  • Tailwind CSS & Radix UI: Modern, accessible, and beautiful
  • State management: React + Next.js context, hooks

Backend & Integrations

  • Prisma ORM: Type-safe database with meeting, participant, and recording models including threading
  • PostgreSQL: Relational database, optimized for connected meeting data
  • AssemblyAI webhook API: Real-time transcription triggers
  • Groq/Llama 3.3: AI summary pipeline integrating context from parent/child meetings
  • Resend: Email invites and notifications
  • Serverless APIs: Next.js API routes, scalable on Vercel/Node

Administrative Features

  • Dashboard: Real-time stats for meetings, recordings, threads
  • Visual Thread Tree: Instantly trace context and evolution of any topic/decision
  • Meeting Management: Create, schedule, invite, record, manage history
  • Role-Based Authorization: Secure admin+user dashboard

Why Mixora Stands Out

  • Only platform with automatic meeting threading and context-aware summaries
  • Transcription & summaries are FREE and UNLIMITED
  • Modern, serverless, and scalable architecture
  • Easy onboarding for teams & organizations
  • Cost-effective: all features included, no per-user licensing
  • Perfect for project management, client comms, compliance, and onboarding

Tags

Next.js, React 19, TypeScript, Prisma, PostgreSQL, AssemblyAI, Groq Llama 3.3, Stream.io, Radix UI, Tailwind CSS, Resend, NextAuth.js


Features

  • Smart video conferencing with automatic AI-powered summaries
  • Industry-first threaded meetings for context continuity
  • Secure, multi-role authentication with JWT and OAuth
  • One-click recording, playback and download
  • Visual dashboard for meetings, threads, and summaries
  • Context-aware, searchable project timelines
  • Free, automatic transcription for every meeting
  • Fully responsive, modern UI/UX

Tech Stack

  • Next.js 16, TypeScript, React 19
  • Tailwind CSS, Radix UI
  • Prisma ORM, PostgreSQL
  • Stream.io Video SDK
  • AssemblyAI (transcription)
  • Groq/Llama 3.3 (summaries)
  • NextAuth.js
  • Resend (email)

This project demonstrates deep technical skills in modern full-stack engineering, cloud SaaS, AI integration, authentication, security, and real-time communications for productivity platforms.