learnright

• 

  home section

• 

  ai chat section

• 

  video and main chat section

Inspiration

YouTube has great educational content, but watching videos is often passive. We wanted to turn the video player into an interactive AI tutor that understands exactly what students are learning.

What it does

LearnRight uses the current video timestamp and transcript context to answer questions, explain difficult concepts, and generate personalized quizzes without interrupting the learning flow.

How we built it

Built with Next.js, TypeScript, the YouTube IFrame API, Gemini models via OpenRouter, and custom transcript extraction and context-selection systems.

Challenges we ran into

We faced challenges with extracting YouTube transcripts, synchronizing the video player with React, and handling incomplete or malformed JSON generated by LLMs.

Accomplishments that we're proud of

We built a functional AI study companion that provides context-aware answers based on the exact section of a video the student is watching.

What we learned

We learned that providing focused context significantly improves AI accuracy, and that robust error handling is essential when building reliable AI applications.

You can replace What's next for LearnRight with:

What's next for LearnRight

We plan to add an interactive assessment mode where students can take tests after completing a video, receive instant feedback, identify knowledge gaps, and get personalized recommendations for what to revise next.

Built With

• ai
• python
• youtube