LearnHub Backend 🚀

Welcome to LearnHub – AI-Powered Personalized Learning Paths
Generate structured, interactive learning paths with AI-generated chapters, curated YouTube videos, and rich study notes – all stored securely on AWS.

---

🔧 What It Does

• 🎯 User inputs a topic, description, category, difficulty, and tone
• 📚 AI (Claude via Amazon Bedrock) generates a complete course with:
  • Chapter titles
  • Learning objectives
  • Key concepts
  • Study notes
  • Practical applications
• 📹 Top YouTube videos are fetched and ranked using view counts, likes, and recency
• 📦 Final course is uploaded to Amazon S3 as a structured JSON
• 🔄 Course is retrievable by ID for rendering on frontend (Explore, Detail pages)

---

🛠 Tech Stack

• Backend: FastAPI (Python 3.9+)
• AI: Amazon Bedrock (Claude 3.5, Titan)
• Video Search: YouTube Data API
• Storage: Amazon S3
• Auth-Ready: Amazon Cognito compatible
• Serverless Support: Mangum for AWS Lambda

---

⚙️ Endpoints

POST /generate-learning-path/

• Accepts course input and returns full structured content
• Includes AI-generated chapters, video links, metadata

POST /upload-course-to-s3

• Uploads a given course to S3
• Returns the s3_uri and generated course_id

GET /get-course/{course_id}

• Fetches course from S3 using the ID

GET /list-courses/

• Lists all saved courses in the S3 folder

---

🚀 Getting Started

1. Clone the repo

git clone https://github.com/your-username/learnhub-backend.git
cd learnhub-backend
---

### 2. Create .env file

AWS_REGION=us-east-1
AWS_ACCESS_KEY_ID=your-access-key
AWS_SECRET_ACCESS_KEY=your-secret-key
S3_BUCKET_NAME=your-s3-bucket
S3_FOLDER=courses
YOUTUBE_API_KEY=your-youtube-api-key

3. Install dependencies
pip install -r requirements.txt

4. Run the server
bash
Copy
Edit
uvicorn main:app --reload

📦 Sample Request
POST /generate-learning-path/
{
  "topic": "Web Accessibility",
  "description": "Learn how to build inclusive web apps",
  "category": "Technical & Programming",
  "difficulty": "Beginner",
  "chapters": 3,
  "tone_output_style": "Educational"
}

🧠 Architecture Highlights
💡 Claude Sonnet 3.5 is used for structured JSON course content

🔁 Retry & throttling logic for Bedrock and YouTube APIs

🧠 YouTube curation uses search + stats API to select top videos

📤 S3 Upload stores each course as courses/{filename}.json

⚡ Mangum used for serverless (AWS Lambda) compatibility

✅ Features Checklist
 Claude prompt-based chapter generation

 YouTube video auto-curation

 Dynamic metadata (chapters, views, timestamps)

 Upload + fetch from S3

 Plug-and-play backend for any frontend

 OpenAPI/Swagger docs enabled by default

💡 Sample Use Cases
🧪 College students wanting bite-sized, YouTube-supported study plans

🧑‍🏫 Teachers generating quick custom curriculum

🎧 Audio learners preparing voice-based lessons (EchoPod integration)

🧩 Plugin for LMS systems to auto-generate paths

🧩 Next Features
 PDF export for full learning path

 Generate quizzes per chapter

 Visual knowledge graphs from chapters

 Course ratings + user feedback collection

 Instructor/Org multi-user support

📁 Project Structure

learnhub-backend/
│
├── main.py               # FastAPI application
├── .env                  # Secrets and keys
├── prompts/              # Prompt building logic
├── models/               # Pydantic schema definitions
├── utils/                # S3, Bedrock, YouTube helpers
├── requirements.txt      # Python dependencies
└── README.md