GrocerAi

Inspiration

• Struggled to find certain items at grocery stores.
• Encountered employees at stores (like Target) who were unsure of item locations, providing only general directions.

What it does

• Web app that creates a personalized shopping list based on user input (e.g., specific events or general shopping needs).
• Identifies healthy options.
• Finds the most efficient route to navigate the grocery store and collect all items.

How we built it

• Used a picture of Kroger’s layout from Google as a reference.
• Physically visited Kroger to map out and document different store sections.
• Integrated the Kroger API to access real-time product data.
• Implemented Dijkstra's algorithm for efficient in-store route navigation.
• Used NLP with LangChain to understand user intent and generate personalized shopping lists.

Challenges we ran into

• Some items on the generated list didn’t make sense or weren’t accurate.
• Faced challenges with state management, which led to changing our tech stack.
• Had difficulty setting up and configuring the server initially.

Accomplishments that we're proud of

• Successfully implemented Dijkstra’s algorithm to optimize in-store navigation routes.
• Converted the grocery store map to a 2D array, labeling different sections using colors.
• Seamlessly integrated the frontend and backend, ensuring smooth functionality.

What we learned

• Learned how to build a full-stack web application from scratch.
• Gained experience with Retrieval-Augmented Generation (RAG) techniques.
• Became proficient in working with multiple APIs.
• Improved our skills in project planning and team collaboration.

What's next for GrocerAi

• Adding social networking features, such as liking items and seeing what friends like.
• Expanding the app to other grocery stores and enhancing functionality.
• Preparing for a launch on the App Store.

Built With

• azure
• css
• html
• javascript
• langchain
• node.js
• python
• react