Skip to content

mrnkim/ShoppableVideo-2

Repository files navigation

Shoppable Video – AI-Powered In-Video Commerce Demo

A sample web application that turns any streaming video into an interactive, “shop-the-look” experience. Built on TwelveLabs’ Analyze API, the app detects products that appear in specific frames, generates contextual descriptions, and lets viewers check-out without pausing playback – mirroring TikTok Shop’s contextual commerce flow but optimized for long-form streaming content.

📌 Demo App – Try the live demo

App screenshot

Video Walkthrough – Watch the demo walkthrough

✨ Key Capabilities

  • Product Detection – AI-powered analysis identifies products in video frames
  • Contextual Product Information – Automatically generates descriptions and product details based on visual content
  • Dynamic Product Markers – Visual indicators appear only when products are detected on screen
  • Amazon Integration – Direct links to Amazon search results for detected products

🛠️ How It Works

App screenshot

  1. Fetch Video – Get the most recent video from the default Twelve Labs index.
  2. Check Metadata – If product metadata exists, display it. If not, proceed to analysis.
  3. Analyze & Save – Analyze the video with Twelve Labs to generate product info, then save it to the video.
  4. Display Products – Show the product details to the user in the UI.

⚙️ Built with

  • Twelve Labs API – For analyzing video and saving analyzed data to each video
  • Next.js 14 / React 18
  • TypeScript
  • Tailwind CSS 3 + MUI
  • React-Player

🚀 Quick Start

1. Clone & install

git clone [email protected]:mrnkim/ShoppableVideo-2.git
npm install

2. Configure environment

Customize .env.example as follows:

TWELVELABS_API_KEY=<YOUR_API_KEY>
NEXT_PUBLIC_DEFAULT_INDEX_ID=<YOUR_DEFAULT_INDEX_ID>
TWELVELABS_API_BASE_URL=https://api.twelvelabs.io/v1.3

3. Run dev server

npm run dev

4. Open in Browser

Open http://localhost:3000

About

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •