VidInsights.ai

Transform YouTube videos into multilingual insights - Summarize, analyze, and interact in your language

Comment

Inspiration

In today's digital age, video content is exploding, but language barriers and time constraints often prevent people from accessing valuable information. We were inspired by the need to make educational YouTube content more accessible to everyone, regardless of their language or time availability. The rise of regional content creators and the growing demand for multilingual accessibility drove us to create a solution that bridges these gaps.

What it Does

VidInsights.ai is a powerful tool that transforms YouTube videos into interactive, multilingual knowledge hubs:

  • Generates concise, customizable summaries in multiple Indian languages (Hindi, Marathi, Gujarati, Bengali, Kannada).
  • Enables natural language Q&A about video content.
  • Supports voice-based questions for enhanced accessibility.
  • Creates knowledge graphs to understand video context and relationships.
  • Offers flexible summary styles (concise, normal, explanatory) and adjustable word counts.
  • Features a modern, intuitive UI with dark/light mode support.

How We Built It

We created VidInsights.ai using a modern tech stack:

  • Frontend: React with Material-UI for a sleek, responsive interface.
  • Backend: FastAPI for efficient API handling.
  • AI
    • LLAMA(Groq) for fast, accurate language processing.
    • Whisper for automatic speech recognition (ASR).
    • Neo4j for knowledge graph creation and querying.
    • YouTube Transcript API for video content extraction.

Challenges We Ran Into

  • Building a robust knowledge graph that captures video relationships effectively.
  • Optimizing API response times for a smooth user experience.
  • Handling voice input processing and transcription reliably.
  • Managing state and caching for better performance.
  • Implementing proper error handling for various API scenarios.

Accomplishments That We're Proud Of

  • Created a fully functional multilingual video insight platform.
  • Successfully implemented a knowledge graph-based Q&A system.
  • Achieved seamless integration of voice input functionality.
  • Built an intuitive, accessible UI that works across devices.
  • Developed efficient caching and history management.

What We Learned

  • Advanced knowledge graph implementation techniques.
  • Efficient state management in React applications.
  • API optimization and error handling strategies.
  • Voice processing and transcription integration.

What's Next for VidInsights.ai

  • Expand language support to include more regional languages.
  • Implement collaborative features for shared learning.
  • Add timestamp-based video navigation from summaries.
  • Enable batch processing of multiple videos.
  • Develop a browser extension for instant access.
  • Add support for more video platforms beyond YouTube.
  • Implement real-time translation during video playback.

Built With

Share this project:

Updates