⚡ Syncora — Enterprise Real-Time Collaboration & AI Workspace OS

1. Overview

Syncora is an enterprise-grade, high-performance real-time collaboration workspace and intelligent team operating system. Designed with an obsidian pitch-black aesthetic and modern glassmorphism, Syncora unifies team chat, multi-party live voice & video meetings, screen sharing, automated meeting intelligence, and task management into a single, cohesive ecosystem.

Real-Time Communication: Instant channel broadcasts and 1-to-1 Direct Messages with typing indicators, read receipts, and live emoji reactions powered by Socket.io.
WebRTC Voice, Video & Screen Sharing: Peer-to-peer 1-to-1 calling and multi-participant channel meetings with ultra-low latency audio, HD video, and screen sharing.
AI-Powered Meeting Intelligence: Automatic generation of comprehensive meeting summaries and word-for-word structured transcripts using cutting-edge LLMs (Gemini / Groq).
Interactive Kanban Board & Task AI: Dynamic drag-and-drop task tracking with conversational AI assistants that query deadlines, summarize workloads, and extract tasks directly from conversation messages.
Multi-Tenant Workspaces & Fine-Grained RBAC: Isolated workspaces, channel permissions, admin moderation, and user bans.
Universal Search & Global Telemetry: Fast full-text search across messages, tasks, channels, and team members with live presence detection.

2. Application Showcase

🌟 1. Channels, Live Meetings & AI Summaries

Dynamic multi-channel team chat with integrated meeting history cards. Access real-time audio/video meetings, launch screen shares, and review AI-generated meeting summaries and transcripts with a single click.

💬 2. Direct Messages, Live 1-to-1 Calls & Chat Management

Private messaging with real-time presence indicators, WhatsApp-style voice and video calling, call log cards, message replies, emoji reactions, and full conversation deletion.

📋 3. Smart Tasks & Syncora Task AI Assistant

Interactive Kanban task board categorized by Pending, In Progress, and Completed states. Features conversational Task AI to query pending tasks, track overdue items, and auto-generate tasks from chat messages.

3. Core Features

💬 Real-Time Messaging & Rich Interactions

Instant Synchronization: Sub-50ms message delivery via optimized Socket.io room routing.
Emoji Reactions & Pinning: WhatsApp-style emoji reaction picker and pinned messages tab for critical announcements.
Threaded Context: Quote-replies with automatic smooth scrolling and visual highlighting to original messages.
Soft Deletes & In-Line Editing: Edit sent messages in place or soft-delete with realtime broadcast updates.
Conversation Deletion: Full chat deletion with confirmation modals that permanently removes chat history across both devices.

🎥 WebRTC Voice, Video & Screen Sharing

1-to-1 Calls: Instant voice and video call signaling with custom ringers, call timers, and fallback states.
Group Channel Meetings: Multi-participant meetings integrated directly into workspace channels with live banner indicators.
Screen Sharing: One-click native desktop screen sharing (getDisplayMedia) with audio streaming.
Call Telemetry & History: Detailed call logs recording duration, status (completed, missed, cancelled, rejected), and quick call-back buttons.

🤖 AI Meeting Intelligence & Assistant Engine

Meeting Summarization: Automatically distills lengthy audio/video discussions into actionable key takeaways, decisions, and action items.
Full Transcripts: Structured, chronological transcript viewer with speaker breakdown and timestamps.
Syncora Task AI: Floating interactive assistant window that answers natural language questions (e.g. "Show my pending tasks", "What tasks are overdue?").
Message-to-Task Extraction: Convert any chat message into a formal task with auto-filled title, description, and source attribution.

📊 Task Management & Kanban Board

Three-Stage Workflow: Seamless status transitions between Pending, In Progress, and Completed.
Priority & Due Dates: High, Medium, Low urgency tags with visual overdue warnings.
Assignee Management: Assign tasks to workspace members with realtime assignment alerts.

🏢 Workspaces & Administrative Control

Multi-Tenant Architecture: Create and switch between multiple workspaces seamlessly.
Role-Based Access: Granular roles (Owner, Admin, Member) controlling channel creation, task assignment, and settings.
User Moderation: Ban disruptive users from specific channels or workspaces with automated permission revocation.

4. System Architecture

graph TD
    classDef client fill:#0B0B0C,stroke:#8B5CF6,stroke-width:2px,color:#F2F0EB;
    classDef gateway fill:#141418,stroke:#6366F1,stroke-width:2px,color:#F2F0EB;
    classDef service fill:#10121A,stroke:#38BDF8,stroke-width:2px,color:#FFF;
    classDef db fill:#0A1918,stroke:#10B981,stroke-width:2px,color:#10B981;
    classDef ai fill:#1A0F1F,stroke:#C084FC,stroke-width:2px,color:#E9D5FF;
    classDef webrtc fill:#1F130B,stroke:#F59E0B,stroke-width:2px,color:#FDE68A;

    A["🖥️ Frontend Client<br/>(React 19 + TypeScript + Vite + Obsidian CSS)"]:::client
    B["⚙️ API Gateway & Auth<br/>(Express.js + JWT Auth Middleware)"]:::gateway
    C["⚡ Real-Time Socket Server<br/>(Socket.io Engine)"]:::gateway
    D["📞 WebRTC Signaling Service<br/>(SDP Offer/Answer & ICE Exchange)"]:::webrtc
    E["🤖 Syncora AI Intelligence<br/>(Gemini 2.5 / Groq LPU Models)"]:::ai
    F["🗄️ MySQL / TiDB Database<br/>(Users, Workspaces, Channels, Messages, Tasks)"]:::db
    G["📁 File Storage Service<br/>(Multer Local & Cloud Storage)"]:::service

    A <-->|REST API Requests| B
    A <-->|Bi-directional WebSockets| C
    A <-->|Peer-to-Peer Audio/Video/Screen| D
    B <-->|Async Queries & Migrations| F
    C <-->|Live State & Presence| F
    B -->|Meeting Summaries & Task AI| E
    B <-->|File Uploads & Attachments| G
    C -.->|Broadcast Notifications| A

5. Database Entity Relationship (ER) Diagram

erDiagram
    USERS ||--o{ WORKSPACE_MEMBERS : "belongs to"
    WORKSPACES ||--o{ WORKSPACE_MEMBERS : "contains"
    WORKSPACES ||--o{ CHANNELS : "owns"
    CHANNELS ||--o{ MESSAGES : "contains"
    USERS ||--o{ MESSAGES : "sends"
    MESSAGES ||--o{ MESSAGE_REACTIONS : "receives"
    USERS ||--o{ MESSAGE_REACTIONS : "reacts"
    USERS ||--o{ DIRECT_MESSAGES : "sends/receives"
    DIRECT_MESSAGES ||--o{ DIRECT_MESSAGE_REACTIONS : "receives"
    USERS ||--o{ DIRECT_MESSAGE_REACTIONS : "reacts"
    WORKSPACES ||--o{ TASKS : "manages"
    USERS ||--o{ TASKS : "assigned_to"
    USERS ||--o{ NOTIFICATIONS : "receives"
    USERS ||--o{ BANNED_USERS : "banned"
    CHANNELS ||--o{ BANNED_USERS : "banned_from"

    USERS {
        int user_id PK
        string name
        string email
        string password_hash
        string avatar_url
        boolean is_online
        datetime last_seen
        datetime created_at
    }

    WORKSPACES {
        int workspace_id PK
        string name
        string description
        int created_by FK
        datetime created_at
    }

    WORKSPACE_MEMBERS {
        int id PK
        int workspace_id FK
        int user_id FK
        string role "owner | admin | member"
        datetime joined_at
    }

    CHANNELS {
        int channel_id PK
        int workspace_id FK
        string name
        string description
        datetime created_at
    }

    MESSAGES {
        int message_id PK
        int channel_id FK
        int sender_id FK
        text message_text
        int reply_to FK
        string file_url
        boolean is_edited
        boolean is_deleted
        boolean is_pinned
        datetime created_at
    }

    DIRECT_MESSAGES {
        int direct_message_id PK
        int sender_id FK
        int receiver_id FK
        text message_text
        int reply_to FK
        string file_url
        string message_type "text | call"
        string call_type "voice | video"
        string call_status "completed | missed | cancelled | rejected"
        int call_duration
        boolean is_read
        datetime created_at
    }

    TASKS {
        int task_id PK
        int workspace_id FK
        string title
        text description
        string status "pending | in_progress | completed"
        string priority "low | medium | high"
        date due_date
        int assigned_to FK
        int created_by FK
        datetime created_at
    }

6. How Real-Time Features Work

📞 1. WebRTC 1-to-1 Calls & Screen Sharing

sequenceDiagram
    autonumber
    actor Caller as Alice (Caller)
    participant Signaling as Socket.io Server
    actor Callee as Bob (Callee)

    Caller->>Signaling: call_user (offer, type: 'video', receiver: Bob)
    Signaling->>Callee: incoming_call (caller: Alice, offer)
    Callee->>Callee: Plays Ringtone & Displays Call Modal
    Callee->>Signaling: answer_call (answer, caller: Alice)
    Signaling->>Caller: call_accepted (answer)
    
    rect rgb(20, 20, 28)
        note over Caller, Callee: Direct Peer-to-Peer (WebRTC) Media Stream Established
        Caller->>Callee: ICE Candidate Exchange (via Socket)
        Callee->>Caller: ICE Candidate Exchange (via Socket)
        Caller->>Callee: Real-Time Audio + HD Video Stream (DTLS / SRTP)
        Callee->>Caller: Real-Time Audio + HD Video Stream (DTLS / SRTP)
    end

    opt Screen Sharing Toggle
        Caller->>Caller: capture getDisplayMedia()
        Caller->>Callee: replaceTrack(videoTrack)
        note over Caller, Callee: Instant Screen Broadcast without Disconnecting
    end

    Caller->>Signaling: end_call (call_duration)
    Signaling->>Callee: call_ended
    Signaling->>Signaling: Record Call History to Database

Signaling Handshake: When Alice clicks "Voice Call" or "Video Call", her browser generates an SDP Offer containing supported codecs and network details. This offer is routed through Socket.io to Bob.
Call Acceptance: Bob receives an incoming_call modal with ringtone audio. Upon clicking accept, Bob's browser creates an SDP Answer and dispatches it back to Alice.
ICE Candidate Exchange: Both peers discover candidate network paths (via STUN servers) and exchange candidates to negotiate the fastest direct peer-to-peer route.
Media Track Swapping (Screen Share): When a user shares their screen, navigator.mediaDevices.getDisplayMedia() acquires the display stream and replaces the existing camera track on the RTCRtpSender seamlessly without restarting the connection.
Call Lifecycle & History: When a call ends or is rejected, duration and statuses are logged into direct_messages as visual call history cards with quick callback shortcuts.

⚡ 2. Channel & DM Real-Time Messaging

Deterministic Room Strategy:
- Channels: All members join room channel_{channelId}. Broadcasts reach only active channel subscribers.
- Direct Messages: Room names are computed deterministically using sorted IDs: dm_{Math.min(userA, userB)}_{Math.max(userA, userB)}. Both users automatically converge on the exact same room ID.
- User Notification Room: Each user joins user_{userId} on connect to receive real-time unread badges and assignment alerts.
Optimistic UI Updates: Reactions, read receipts, and message state updates render immediately in the client UI and synchronize across peers in under 50ms.

🤖 3. AI Meeting Summaries & Transcripts

Audio / Context Processing: During live meetings, discussion points, speaker metadata, and meeting duration are captured.
LLM Synthesis: The meeting context is passed through Gemini / Groq models with tailored prompt templates for:
- Executive Summary: Core highlights and purpose.
- Key Decisions: Bullet points of agreements made.
- Action Items: Concrete task assignments.
- Full Transcript: Chronological speaker breakdown.
Interactive Modal: The generated summaries and transcripts are rendered using formatted markdown and syntax highlighting directly in the channel.

7. Technical Stack

Layer	Technology	Key Capabilities & Purpose
Frontend UI	React 19, TypeScript, Vite	Ultra-fast client, strict type safety, modular architecture
Styling & Theme	Obsidian Dark Theme (Custom CSS)	Pitch black palette, glassmorphism, responsive mobile flexbox
Icons & Media	Lucide React	Clean, scalable vector iconography
Backend Framework	Node.js, Express.js	High-throughput REST API routing and middleware pipelines
Real-Time Engine	Socket.io 4.x	Low-latency WebSockets, room management, presence detection
Audio / Video Mesh	WebRTC Native API	Peer-to-peer audio, video calling, and live screen sharing
Database	MySQL 8.0 / TiDB Serverless	ACID relational storage, foreign key cascades, migration tooling
AI LLM Inference	Google Gemini 2.5 & Groq LPU	Meeting summarization, transcript generation, conversational Task AI
Authentication	JWT (JSON Web Tokens) & BCrypt	Stateless bearer authentication and secure password hashing
File Handling	Multer	Multi-format file attachments with size and type validation

8. Local Setup Guide

Prerequisites

Node.js (v18.x or higher)
npm or yarn
MySQL or TiDB instance

1. Clone the Repository

git clone https://github.com/gpranit16/syncora.git
cd syncora

2. Configure Database & Migrations

Create a MySQL / TiDB database:

CREATE DATABASE smart_team_collab CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;

Import the database schema:

mysql -u root -p smart_team_collab < database/schema.sql

3. Setup and Run Backend

cd backend

# Install dependencies
npm install

# Create environment configuration (.env)
cat <<EOF > .env
PORT=5000
DB_HOST=localhost
DB_USER=root
DB_PASSWORD=your_mysql_password
DB_NAME=smart_team_collab
JWT_SECRET=your_jwt_secret_key_here
GEMINI_API_KEY=your_gemini_api_key_here
EOF

# Start backend server
npm run dev

The backend server will launch on http://localhost:5000.

4. Setup and Run Frontend

cd ../frontend

# Install dependencies
npm install

# Start Vite development server
npm run dev

Open your browser and navigate to http://localhost:5173.

9. API Reference Overview

Endpoint	Method	Description
`/api/auth/register`	`POST`	Register a new user account
`/api/auth/login`	`POST`	Authenticate user and obtain JWT token
`/api/workspaces/my-workspaces`	`GET`	List all workspaces for authenticated user
`/api/channels/workspace/:id`	`GET`	Get all channels belonging to a workspace
`/api/messages/channel/:id`	`GET`	Retrieve chat history for a channel
`/api/direct-messages/chat/:id`	`GET`	Get DM conversation with specific user
`/api/direct-messages/conversation/:id`	`DELETE`	Delete full DM conversation with user
`/api/tasks/workspace/:id`	`GET`	Get all Kanban tasks for a workspace
`/api/tasks/create`	`POST`	Create a new task with priority and assignee
`/api/ai/meeting-summary`	`POST`	Generate AI summary for channel meeting
`/api/search?q=query`	`GET`	Global search across messages, tasks, and users

10. License

Distributed under the MIT License. See LICENSE for more information.

_{Built with ❤️ by Pranit Kumar}

Name		Name	Last commit message	Last commit date
assets/screenshots		assets/screenshots
backend		backend
config		config
database		database
docs		docs
frontend		frontend
.gitignore		.gitignore
README.md		README.md
standalone_deepgram_test.js		standalone_deepgram_test.js

Navigation Menu

Repository navigation

About

Repository files navigation

⚡ Syncora — Enterprise Real-Time Collaboration & AI Workspace OS

1. Overview

2. Application Showcase

🌟 1. Channels, Live Meetings & AI Summaries

💬 2. Direct Messages, Live 1-to-1 Calls & Chat Management

📋 3. Smart Tasks & Syncora Task AI Assistant

3. Core Features

💬 Real-Time Messaging & Rich Interactions

🎥 WebRTC Voice, Video & Screen Sharing

🤖 AI Meeting Intelligence & Assistant Engine

📊 Task Management & Kanban Board

🏢 Workspaces & Administrative Control

4. System Architecture

5. Database Entity Relationship (ER) Diagram

6. How Real-Time Features Work

📞 1. WebRTC 1-to-1 Calls & Screen Sharing

⚡ 2. Channel & DM Real-Time Messaging

🤖 3. AI Meeting Summaries & Transcripts

7. Technical Stack

8. Local Setup Guide

Prerequisites

1. Clone the Repository

2. Configure Database & Migrations

3. Setup and Run Backend

4. Setup and Run Frontend

9. API Reference Overview

10. License

About

Releases

Packages

Contributors

Languages

Latest commit

History

Folders and files

Repository files navigation

⚡ Syncora — Enterprise Real-Time Collaboration & AI Workspace OS

1. Overview

2. Application Showcase

🌟 1. Channels, Live Meetings & AI Summaries

💬 2. Direct Messages, Live 1-to-1 Calls & Chat Management

📋 3. Smart Tasks & Syncora Task AI Assistant

3. Core Features

💬 Real-Time Messaging & Rich Interactions

🎥 WebRTC Voice, Video & Screen Sharing

🤖 AI Meeting Intelligence & Assistant Engine

📊 Task Management & Kanban Board

🏢 Workspaces & Administrative Control

4. System Architecture

5. Database Entity Relationship (ER) Diagram

6. How Real-Time Features Work

📞 1. WebRTC 1-to-1 Calls & Screen Sharing

⚡ 2. Channel & DM Real-Time Messaging

🤖 3. AI Meeting Summaries & Transcripts

7. Technical Stack

8. Local Setup Guide

Prerequisites

1. Clone the Repository

2. Configure Database & Migrations

3. Setup and Run Backend

4. Setup and Run Frontend

9. API Reference Overview

10. License

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages