PantryPilot

• 

  The user's current pantry.

• 

  Our loading screen while Gemini AI parses the receipt.

• 

  Suggested recipes given to the user after they scan their receipt.

• 

  Our original app flow brainstorm.

Inspiration

Keeping track of your groceries can be very difficult with how overwhelming life can be. Current solutions often require you to tediously scan every item in your pantry, and manually input ones without barcodes! What if there was a solution that allows you to take note of everything in your pantry as you purchase it, suggest recipes that add a pleasant twist to your favourite dishes, and save money by minimizing food waste? PantryPilot lets home dining take flight!

What it does

PantryPilot scans your receipts from the grocery store. Using OCR and an LLM, it parses this data into a JSON file, which then gets matched with recipes based on those with the best match to the ingredients in your pantry.

How we built it

We built it using various frameworks integrated together, incuding Node.js, Flutterflow, Gemini, Express.js, Tesseract, and Firebase.

Challenges we ran into

With our backend Node.js server and frontend FlutterFlow app both working flawlessly on their own, troubles came when we tried to integrate the two. Through careful debugging and some compromise, we were able to overcome these obstacles.

Built With

• express.js
• firebase
• flutterflow
• gemini
• javascript
• node.js
• tesseract