WhirlWain - Smart Shopping Assistant

"List, Click, Shop!"

Comment

🛒 Smart Shopping Assistant

Smart Shopping Assistant is a cloud-powered web application that helps users effortlessly generate grocery lists either by uploading handwritten notes or by simply describing what they want to cook. Using AWS services and AI, the app generates a clean, actionable list — and lets users directly add products to the cart from one place.

💡 Inspiration

We often forget items while shopping or struggle with messy handwritten lists. With the rise of AI and cloud services, we thought:
What if creating grocery lists could be automated and intelligent?

That’s how the idea of Smart Shopping Assistant was born — an app that turns images or plain English into a clean, ready-to-use shopping list.

We were also slightly inspired by Amazon Rufus — the AI assistant that helps users shop smarter. Our goal was to build something similar, but more personal and versatile for grocery planning, powered by serverless architecture and public AWS tools.

🎯 Most importantly, we wanted to save users the hassle of visiting each product page individually to add items to the cart. Our assistant brings all suggested products into one place — making it faster and easier to add them directly.

🧠 What We Learned

Throughout the project, we explored and applied:

  • AWS Textract for Optical Character Recognition (OCR) and text extraction from images
  • AWS Bedrock (Llama 3) for understanding user scenarios and generating smart grocery lists
  • Amazon Product APIs to fetch real-time item suggestions
  • Hosting scalable frontends using AWS Amplify
  • Designing serverless, cloud-native APIs with AWS API Gateway and Lambda
  • Writing clean, modular React.js code for a smooth and responsive user interface

🔧 How We Built It

  • Frontend: React.js hosted on AWS Amplify
  • OCR: AWS Textract extracts text from uploaded grocery list images
  • AI List Generation: AWS Bedrock (Llama 3) interprets user input and generates grocery lists
  • Backend: AWS Lambda functions handle OCR logic, AI prompts, and product integration
  • API Gateway: Manages requests between the frontend and backend
  • Product Suggestions: Amazon Product APIs deliver relevant, real-time shopping items
  • Version Control: Git & GitHub for collaboration and project tracking

⚠️ Challenges We Faced

  • Prompt engineering: Getting structured and relevant output from LLMs required extensive tuning and testing
  • Service integration: Combining multiple AWS services into a smooth, secure workflow (Textract → Bedrock → Product API) involved learning a lot on the go
  • Time constraints: Working under tight timelines, especially balancing frontend polish with backend functionality, was a constant push

🚀 Conclusion

This project gave us hands-on experience with real-world cloud-based AI application development.

We’re proud of the result — a smart assistant that simplifies everyday shopping with intuitive user interactions and powerful backend intelligence.

✅ One of our main goals was to eliminate the need for users to visit each product page individually to add items to the cart. Instead, our app allows them to add all suggested products directly from one place — fast, simple, and efficient.

Looking ahead, we plan to explore exciting features like:

  • 🗣️ Voice-based list generation
  • 🌐 Multilingual support
  • 💰 Budget-based suggestions
  • 📱 A mobile-first version

We’re excited to keep improving and evolving this assistant to make smart shopping accessible to everyone!

Built With

Share this project:

Updates