Re:Home

Inspiration

In major cities like Seattle, finding shelters nearby and food banks can be challenging, especially for individuals facing homelessness. Our inspiration came from wanting to create an accessible tool that helps people locate shelters near them, while also helping the shelters manage resources more efficiently.

What it does

Platform that allows shelters, people in need, and donors to register themselves.
Individuals in need can quickly find nearby resources using our interactive map-based interface
Let’s volunteers sign up and get in touch with organizations
Displays distance and travel time from a user’s current location to their desired shelter
Tracks inventory movement ## How we built it We built this platform using a myriad of tools such as python, webscraping, Gemini API, streamlit, pytorch, and machine learning. We used a modular development system to break down our whole full-stack website in to smaller functional pieces, then integrating them all. ## Challenges we ran into We broke the project down into multiple subsections. Integrating all the subsections together was one of the biggest challenges we ran into. We faced problems with testing due to a lack of data from users, due to privacy concerns with webscraping, so we had to create synthetic data to make our website functional. We had to use a complex ML prediction algorithm, which was difficult to implement. ## Accomplishments that we're proud of We're proud of being able to put together a resource that can create a positive impact tangibly, connecting those in need to those who can help. We're proud of being able to complete a full-stack development process, implementing a cohesive front and back end in a limited time frame. We're proud of being able to learn and implement new API's and technical skills in an effective manner. ## What we learned We learned how to use the Gemini API with Python to have LLM functionality in our project, we learned how to implement interactive map tools into our Streamlit website, and we learned how to use Temporal Convolutional Networks with PyTorch to make a predictive algorithm for inventory management. We learned how to webscrape lists of websites to get important data and store it in a database. ## What's next for Re:Home We aim to implement a strong encryption algorithm to protect user/organization information in order to publicize our product. We also hope to generate a large user infrastructure to make Re:Home an effective tool in getting people what they need. We hope to launch this tool in a localized, smaller area, such as Seattle to test effectiveness, then later nationalize the project and make a larger global impact.