Inspiration

In the U.S., everyday grocery prices are rising and affordability is a growing pain point. Studies show that buying in bulk could cut costs by about 27 % on average compared to buying smaller quantities, saving households hundreds of dollars a year if they could unlock wholesale pricing at scale.

What it does

UnionBuy makes groceries more affordable by making bulk products available as individually sold products for those unable to afford the entire bulk package.

How we built it

Utilizing agentic AI from AgentCore by AWS, we were able to turn messy and complicated streams of product data into structured outputs that other agents were then able to utilize to uncover hidden savings for the community,

Challenges we ran into

With all the different sources of data available (all in different formats as well), we were faced with the challenge of figuring out how to effectively funnel all this data into one cohesive result. For example, when figuring out how to determine the retail price, we were faced with options from many different smaller grocery stores such as Whole Foods, Erewhon, etc.

Accomplishments that we're proud of

Being able to put everything together (product listings, bulk thresholds, pricing, route mapping, etc.) in such a short amount of time. Our communication as a team really helped us when it came to putting together so many moving parts and ensuring that the agent outputs, product listings, and route mapping worked end to end.

What we learned

The less hard coding we had to do, the less problems we ran into down the line. For example, instead of hard coding in regex patterns and scraping to extract information, we used agents to semantically extract product information.

What's next for UnionBuy

Share this project:

Updates