Clipmark is a high-performance, 100% native YouTube processing tool that transforms any video into navigable timestamps, detailed transcripts, and pro-level SEO metadata. By bypassing AI latency, Clipmark provides near-instant results directly from official YouTube data.

Tip

New to the project? Check out our Full Technical Documentation for a deep-dive into every step, workflow, and API logic.

---

✨ Key Features

• 📊 Comprehensive Video Insights: Real-time extraction of view counts, likes, comments, and channel metadata.
• 🔖 Native Chapter Parsing: Automatically extracts "creator-intended" chapters from video descriptions with click-to-seek functionality.
• 📝 Optimized Transcripts: Word-for-word YouTube captions grouped into 10-second readable segments.
• 🔍 SEO Metadata: Instant retrieval of full video descriptions and tags for SEO optimization.
• ♾️ 100% Free & Unlimited: No video processing limits or tier-based restrictions.
• ⚡ Lightning Fast: Native-only processing path skips heavy audio extraction and STT for near-instant results.

---

🛠 Tech Stack

• Frontend: Next.js 14 (App Router), Tailwind CSS, Clerk Auth, Lucide React.
• Backend: Node.js (Express), BullMQ (Redis Queue), Mongoose (MongoDB).
• External APIs: YouTube Data API V3, YouTube Transcript API.
• Storage/Exports: Cloudinary (for file exports).

---

🚀 Getting Started

1. Prerequisites

Ensure you have the following installed:

• Node.js (v18 or higher)
• Redis (for the BullMQ worker)
• MongoDB (local or Altas)

2. Environment Setup

Create a .env file in apps/api and a .env.local in apps/web.

Apps API (apps/api/.env)

PORT=8080
MONGODB_URI=your_mongodb_uri
REDIS_URL=redis://localhost:6379
YOUTUBE_API_KEY=your_google_api_key
CLERK_SECRET_KEY=your_clerk_secret_key
CLERK_PUBLISHABLE_KEY=your_clerk_publishable_key
CLOUDINARY_URL=your_cloudinary_url

Apps Web (apps/web/.env.local)

NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY=your_clerk_publishable_key
CLERK_SECRET_KEY=your_clerk_secret_key
NEXT_PUBLIC_API_URL=http://localhost:8080

3. Installation

Install dependencies for the entire monorepo:

npm install

4. Running the Project

Start both the Frontend and Backend concurrently:

npm run dev

The apps will be available at:

• Web: http://localhost:3000
• API: http://localhost:8080

---

📂 Project Structure

├── apps/
│   ├── api/            # Express Backend (Jobs, Workers, API)
│   └── web/            # Next.js Frontend (Dashboard, Results, Auth)
├── packages/
│   └── shared/         # Shared Types and Constants
└── package.json        # Root workspace configuration

---

🔄 Workflow logic

1. Submission: User pastes a YouTube URL into the dashboard.
2. Metadata Fetch: Backend uses YouTube Data API to fetch title, duration, and detailed stats (views, likes).
3. Queueing: A job is added to BullMQ for processing.
4. Native Processing:
   • Extraction of original description.
   • Regex parsing of description for timestamps (Chapters).
   • Retrieval of official closed captions.
   • Grouping transcript into 10s intervals.
5. Database Update: Results are stored in MongoDB.
6. Delivery: Frontend polls for completion and displays the interactive dashboard with Click-to-Seek.

---

📄 License

© 2026 Clipmark. All rights reserved. Built by Jenil Rupapara.