Supplyspring

Inspiration

Wellspring Women's Center supports women and children in need every day, but donation tracking and inventory management were still being handled manually. We wanted to create a centralized system that could save staff time, reduce errors, and make it easier to see incoming and outgoing donations in real time.

What it does

Our platform acts as a real-time donation and inventory management hub for Wellspring Women's Center. Staff members can track donated items, monitor inventory levels, view trends through dashboards, and receive live updates whenever items are added or distributed.

How we built it

We used Next.js, TypeScript, React, and Prisma for the database. We used OpenAI vision and language models to help implement both the voice and photo features for the logging logic, while also including a step where these tools are checked by the user working with them. We used both OpenAI and ElevenLabs API's for this project as well.

Challenges we ran into

The biggest struggle we faced was being able to balance accuracy with automation. Since donation centers don't blindly accept every donation, our AI implementations need to be reviewed and checked against Wellspring's acceptance rules. Problems arose when handling rejected items and preventing valid donations from also being rejected.

Accomplishments that we're proud of

We’re proud of a few of the features that we created for the platform with a direct real-world impact for Wellspring Women's Center. We implemented both a voice and picture logging feature. The voice feature allows the worker to say all the items that have been donated, then allows the worker to check to see if it's correct/accurate and log the items. The photo feature allows the worker to take a picture of inventory, and an LLM has been trained to recognize what the item is and the count of the item; these pictures can have multiple different items in them. The dashboard provides clear and accessible data visualization, and the live inventory updates make tracking donations significantly more efficient than manual methods. We’re also proud of building a polished and scalable interface within a limited hackathon timeframe.

What we learned

We learned how important non-profits are to usability and accessibility. Fancy features will be mostly useless in these types of projects unless they are extremely user-friendly and easy to learn. We also learned how AI can be used to assist, and we had to manage this assistance so as not to replace judgment and keep that to the user.

Through this project, we learned more about designing software for nonprofits and real users rather than just technical demos. We also gained experience with real-time state management, frontend data visualization, and building user-focused dashboards that prioritize simplicity and accessibility.

What's next for Wellspring Women's Center

Next, we would improve the role-access for staff/volunteers. There should be a stronger audit history implemented in the future. An additional feature that would make this tool a lot easier to use would be barcode support.