Book Tracker: AWS Serverless App

✨ Overview

This is a fully serverless application that allows users to track books they are currently reading, have read, or want to read. The application uses modern AWS serverless technologies including API Gateway, Lambda, DynamoDB, Cognito, S3 (with signed URLs), and optionally AWS Bedrock for AI-powered recommendations.

The goal is to provide a complete reference implementation for developers looking to learn and deploy real-world serverless applications on AWS.

📆 Features

Core

• User sign-up and authentication via Amazon Cognito

• Track books by status:

  • Want to Read
  • Currently Reading
  • Finished Reading

• Store book metadata: title, author, status, rating, review, tags

• Allow users to update or delete book entries

Export & Import (S3 Signed URLs)

• Export: Generate downloadable reading summaries in CSV, JSON, or PDF format
• Import: Import complete book collections from JSON exports with 100% fidelity
• Files stored in S3 with private permissions and signed URLs for secure download/upload

Book Search

• Search for books using Google Books API (title, author, etc.)
• Auto-fill book metadata when adding a new book
• Fallback to manual entry if desired

AI Integration (AWS Bedrock)

• Use Bedrock (e.g. Claude or Titan) to generate personalized book recommendations
• Prompt based on user's reading history and preferences
• Return results with titles, authors, and genres

Optional Enhancements

• Audio review uploads to S3 (voice memos)
• Periodic email report with reading history (via SES)
• Social features: shared book lists, reading challenges

🛁 Architecture

Overview

Frontend (React/S3/CloudFront)
    |
    |---> API Gateway (HTTP API + JWT Authorizer)
             |
             |---> Lambda Functions
             |        |- Book CRUD
             |        |- Book Search Proxy (Google Books)
             |        |- Recommendation Engine (Bedrock)
             |        |- Export Generator (S3 Signed URL)
             |        |- Import (JSON)
             |
             |---> DynamoDB (Books Table)
             |---> S3 (Reports Bucket)
             |---> Cognito (User Auth)

📊 Data Model

DynamoDB: Books Table

Partition Key: PK = USER#<user_id>
Sort Key: SK = BOOK#<book_id>

Attributes:

{
  "title": "The Hobbit",
  "author": "J.R.R. Tolkien",
  "status": "reading",
  "rating": 5,
  "review": "Epic fantasy classic",
  "tags": ["fantasy", "classic"],
  "started_at": "2025-06-01",
  "finished_at": null
}

GSI for Status-Based Queries

• GSI1 Partition Key: status
• GSI1 Sort Key: PK

🔧 API Endpoints

Book Management

GET    /books              --> List all books
GET    /books?status=...  --> Filter by status
POST   /books              --> Create new book
PUT    /books/{id}         --> Update book
DELETE /books/{id}         --> Delete book

Export & Import

POST   /export             --> Generate export file (CSV/JSON/PDF), return signed S3 URL
POST   /import             --> Import from JSON file
GET    /import/{import_id} --> Check import status and results

Search

GET    /search?q=the+hobbit   --> Proxy to Google Books API, return suggestions

Recommendations

GET    /recommendations       --> Bedrock prompt based on reading history

🛡 IAM & Security

• All API endpoints protected via Cognito JWT authorizer
• Users only access their own books via IAM policies + DynamoDB conditions
• S3 bucket private, files accessed via short-lived signed URLs
• Lambda has scoped-down permissions (DynamoDB, S3, Bedrock)

🚀 Deployment

Recommended: AWS SAM or AWS CDK

• sam build && sam deploy for end-to-end setup
• Use environment variables for API keys (e.g. Google Books)

📁 Directories & Files

/infra               --> CDK or SAM templates
/backend             --> Lambda handlers (Python or Node.js)
/frontend            --> React app (optional)
/docs                --> Diagrams, design notes
/tests               --> Unit and integration tests

📖 External APIs

Google Books API

• Endpoint: https://www.googleapis.com/books/v1/volumes?q=...
• No auth needed for basic usage
• Rate-limiting may apply; use backend proxy if needed

AWS Bedrock

• Use Claude or Titan via bedrock:InvokeModel

• Construct a prompt like:

  A user has read: Dune, The Martian. Recommend 5 similar books with genre.

📥 Import & Export Specifications

Import (JSON)

Format: JSON array of book objects with complete schema validation

Example:

[
  {
    "title": "The Hobbit",
    "author": "J.R.R. Tolkien",
    "series": "Middle-earth",
    "status": "read",
    "rating": 5,
    "review": "Classic fantasy adventure",
    "tags": ["fantasy", "classic"],
    "started_at": "2024-01-15",
    "finished_at": "2024-02-01",
    "thumbnail": "https://...",
    "type": "Fantasy",
    "comments": "Perfect introduction to Middle-earth"
  }
]

Validation Rules:

• Required fields: title, author, status
• Status must be: read, currently-reading, want-to-read
• Rating: Integer 1-5 or null
• Dates: ISO format (YYYY-MM-DD) or empty
• Duplicate handling: Skip books with existing IDs

Export Formats

JSON: Complete data preservation with all fields
CSV: Tabular format for spreadsheet import
PDF: Formatted reading report (future enhancement)

✅ Future Enhancements

• Import/Export:
  • PDF export generation (HTML-to-PDF via Puppeteer)
  • Import from other platforms (Goodreads, Amazon)
  • Bulk edit imported books
• Collaboration:
  • Book club / shared lists between users
  • Reading challenges and social features
• Automation:
  • SES email report delivery
  • Scheduled backup exports
• Mobile: React Native or Flutter frontend
• Tech Debt:
  • Avoid hard-coding endpoints
  • Common JWT library/logic
  • Cache-control headers on S3/API

🙏 Contributions

This is intended as a hands-on learning project. Contributions welcome via pull request or issue.

⚡ Credits

Created as a learning tool to explore AWS Serverless, Bedrock, and best practices in full-stack development.

✨ License

MIT License (or choose your own)