Cloud Computing · intermediate

Serverless AI Content Summarizer with Amazon Bedrock

A fully serverless web app where users upload docs for AI summarization using AWS Lambda and Bedrock (Claude 3).

Project overview

Employers want engineers who can integrate LLMs into cloud infrastructure without managing servers or GPUs — this is the modern way.

Real-world use case

A news aggregator platform providing quick summaries of thousands of daily articles for mobile users.

Tools and technologies

• AWS Lambda
• Amazon Bedrock
• S3
• DynamoDB
• SNS

Skills you can demonstrate

Serverless, GenAI Integration, Event-Driven

Build roadmap

1. Enable Bedrock in your AWS account
   Request access to Claude 3 models in the Bedrock console. This can take a few hours to approve.
2. Build the upload flow
   S3 bucket for uploads. Pre-signed URLs so the frontend can upload directly without going through a server.
3. Write the Lambda function
   S3 trigger fires on upload. Lambda extracts text (PyPDF2 for PDFs, just read for text files) and calls Bedrock.
4. Store results in DynamoDB
   Schema: id, filename, original_text_length, summary, created_at. Let the frontend query by id.
5. Build the React frontend
   Upload page, status tracking, summary display. Host on S3 + CloudFront for pennies per month.
6. Add SNS notifications
   When a summary is ready, push to SNS so users get an email. Document the cost per summary vs. hosting Llama on an EC2 GPU.

Ways to extend the project

Add an AI-powered "Sentiment Analysis" feature that flags the tone of the document alongside the summary.

How to showcase your work

• GitHub: Full serverless architecture with IaC templates.
• LinkedIn: Highlight 90% cost reduction vs. traditional server-based AI hosting.
• Resume: Focus on serverless + GenAI integration.