Home Original page

StreamStudy

Inspiration

In today’s digital age, learners face scattered resources, rigid formats like PDFs or long videos, and limited access to adaptive learning. We wanted to create a solution that combines structure, personalization, and accessibility. Inspired by the growing need for on-the-go education, we envisioned StreamStudy — a hybrid AI-powered learning platform for audio and video learning.

What it does

StreamStudy allows users to:

  • Input any topic
  • Choose category, difficulty, tone, and voice
  • Generate a personalized course in audio or video format

For audio:

  • AI generates intro, chapters, and transcripts
  • Converts content into lifelike speech using Amazon Polly

For video:

  • Uses YouTube API to fetch relevant videos
  • Generates chapter-wise learning paths and notes using Claude AI

All content is delivered in-app with real-time generation updates and full playback control.

How we built it

Frontend:

  • Built using Flutter for cross-platform delivery
  • Integrated audio playback with controls (speed, volume, sleep timer)
  • Managed global audio state using Provider
  • WebSocket support for real-time generation status

Backend & Cloud Stack:

  • Amazon Bedrock + Claude: Content generation (text + notes)
  • Amazon Polly: Text-to-speech for podcast-style narration
  • AWS Lambda: Stateless backend functions
  • API Gateway: REST and WebSocket API endpoints
  • Amazon S3: Stores transcripts and audio files
  • Amazon DynamoDB: Metadata and topic storage
  • Amazon Cognito: Secure user authentication
  • YouTube Data API: Fetches and ranks videos
  • Amplify Flutter: Connects Flutter app with AWS services

Challenges we ran into

  • Handling multi-stage real-time updates via WebSocket
  • Syncing transcript display with generated audio
  • Parsing and ranking noisy video data from YouTube
  • Ensuring AWS Amplify configuration works smoothly across platforms
  • Balancing fast generation with high content quality

Accomplishments that we're proud of

  • Seamlessly combined AI-generated audio and video learning into a single platform
  • Built a clean, intuitive UI for both podcast-style and video courses
  • Enabled real-time generation feedback through WebSockets
  • Created dynamic learning paths based on user input and content quality

What we learned

  • Working with generative AI tools like Claude requires thoughtful prompting and output validation
  • Flutter + AWS Amplify is powerful but needs careful environment setup
  • Managing audio state globally and transitioning across screens was key to a smooth UX
  • Real-time app feedback drastically improves user trust and engagement

What's next for StreamStudy

  • Add offline download support for audio courses
  • Enable bookmarking, history, and saved progress
  • Create collaborative playlists and social sharing
  • Build an admin dashboard for moderation and analytics
  • Introduce user feedback and rating for course quality

Updates