Inspiration
Online shopping on modern marketplaces is flooded with thousands of products, mixed reviews, and price fluctuations. Consumers often struggle to verify if a deal is genuine or if product reviews are biased. We built AmzGuard AI to bring an intelligent, trusted verification agent directly into the shopping flow, making e-commerce safer and more personalized.
What it does
AmzGuard AI acts as a smart retail co-pilot:
- Review Authenticity & Sentiment Analysis: Evaluates customer reviews to detect fraudulent or sponsored feedback.
- Smart Product Comparison: Compares product features, prices, and specs dynamically based on user preferences.
- Budget & Safety Guard: Ensures secure checkout recommendations and budget limits before processing orders. ## How we built it
- AI Core: Built using Amazon Bedrock / AWS Claude API to handle intent parsing and sentiment analysis.
- Backend: Python microservice running on AWS Lambda / Render.
- Frontend/Interface: Web application integration with Amazon Developer tools and REST APIs.
Challenges we ran into
Integrating real-time sentiment extraction while maintaining fast response times required optimizing prompt structures and streaming API responses efficiently.
Accomplishments that we're proud of
Successfully building an end-to-end conversational shopping assistant that provides instant risk scores and concise summaries for complex product catalogs.
What we learned
We learned how to leverage Amazon Web Services (AWS) AI infrastructure to process customer sentiment, parse complex product catalogs, and deliver real-time risk scores for e-commerce security.
What's next for AmzGuard AI
Adding voice interaction support, multi-currency support, and expanded integration across various e-commerce platforms.
Log in or sign up for Devpost to join the conversation.