🛠️ Lab 7 — Smart Document Analyzer

Type: Guided lab Time: ~90 minutes Difficulty: Intermediate Cost: Free (Textract: 1000 free pages/month for 3 months)

🎯 What you’ll build

Upload any PDF or document image to S3 and watch AWS automatically extract the text (Textract), analyze the sentiment + entities (Comprehend), and save the results to DynamoDB. Real production AI in 90 minutes.

📋 Prerequisites

  • Comfortable creating Lambda functions
  • Basic Python (provided starter handles 90% of it)
  • An understanding of IAM roles for Lambda

🛠️ AWS Services used

  • S3 — document uploads
  • Lambda — orchestrator that chains the AI calls
  • Textract — pulls text from PDFs and images
  • Comprehend — sentiment + entity extraction
  • DynamoDB — stores analysis results

📖 The full walkthrough

Open Week4_Lab7_Smart_Doc_Analyzer.html in your browser. It’s a complete step-by-step guide with screenshots, code snippets, and inline hints.

💡 The HTML file is self-contained — no install, no internet needed once you’ve cloned the repo. Just double-click to open in your browser.

🆘 If you get stuck

  1. Re-read the relevant step in the HTML walkthrough
  2. Check the Common Mistakes callouts (red boxes in the walkthrough)
  3. Expand the 💡 Hint sections for guidance
  4. WhatsApp the group with a screenshot of the error
  5. Tag your post with #week4done

🎓 Submission

  1. Screenshot of a DynamoDB item showing extracted text + sentiment + entities from a real document
  2. Bonus: try it with your own CV and screenshot the entities Comprehend pulled out

⬅️ Back to Week 4