Inspiration
People often know their budget, intended use, and preferred specifications without knowing which product to choose. ShopMate helps turn that shopping brief into a shortlist and recommendation.
What it does
ShopMate uses Gemini to extract product requirements from a natural-language request, filters its sample catalogue, and recommends a match with an explanation. Users can also search Channel3 for live product listings and retailer offers. USD offers show an approximate INR equivalent while keeping the original price visible. Users can add sample-catalogue items to a cart and try a PayPal Sandbox checkout.
How we built it
The app is built with Python and Streamlit. Gemini extracts the requirements and recommends from the matching catalogue items. Channel3 provides live product search and retailer offers. Frankfurter provides a daily USD/INR reference rate for approximate price display. PayPal’s Orders API creates the fixed-value Sandbox payment; the app verifies the returned order before capture.
Challenges we ran into
Channel3’s current default market is US/USD, while the sample catalogue uses INR. ShopMate shows an approximate rupee conversion for comparison and preserves the retailer’s original USD price. Shipping, taxes, and final retailer prices may differ.
Accomplishments that we're proud of
We brought AI recommendations, live product discovery, sample-cart management, and PayPal Sandbox order verification into one shopping experience. The app checks that an AI recommendation matches a product in its results and verifies the PayPal order before capturing payment.
What we learned
AI output needs validation before it drives product choices, payment flows need explicit order verification, and currency conversions should be clearly labeled as estimates.
What's next for ShopMate AI
Add supported India-market product sources, connect live listings to a clearly priced checkout flow, and use durable storage for pending orders and receipts.
