Price Pilot

Project Presentation Link

link

Inspiration

Households spend a lot of time and money shopping without knowing which stores have the best prices. They may not have the time nor will look at many different stores to see where their groceries are the cheapest. We wanted to help families save money and effort using the power of artificial intelligence.

What it does

Price Pilot tracks past purchases and some metadata, estimates which stores are cheapest for your items.

How we built it

We used JavaScript and Supabase for the backend, HTML/CSS for the interface, and the Claude and Google APIs to estimate prices and make predictions on price and restock timings. We use node.js in order to download certain Supabase functions for Claude Code.

Challenges we ran into

We were first going to create our own web scraper, but we ran into a CORS issue. Many websites do not allow surface-level web scraping for security reasons, so we pivoted to using the Claude API to find and/or estimate prices.

Accomplishments that we're proud of

We successfully created an account system for users to track their purchase history. We were able to analyze the price differences between stores, and display summaries of the info.

What we learned

We learned how to properly vibe code and got firsthand experience in integrating software and knowledge about APIs.

What's next for Price Pilot

Next, we plan to create a mobile app for use on the go, implement receipt scanning so users can simply scan their receipt (which we can use convolutional neural networks for) and track other purchase metadata. We also plan to implement a notification system to users when we predict that they will need a product using time series analysis and their past purchases.

Built With

• claudeapi
• claudecode
• cursor
• google
• html
• javascript
• liveserver
• node.js
• supabase