sunryse

Sunryse - Detect. Predict. Prevent Homelessness

Sunryse

Detect. Predict. Prevent Homelessness

Introduction

Sunryse is HMIS system that could potentially help caseworkers intervene before families end up on the street.

Part 1: The Model

Apart from CSV data provided by GlobalHackthon, we used data from BLS.gov, Data.gov, Numbeo, etc to build clusters of homeless profile. These clusters will help us run it against the incoming data streams from various agencies and companies like, Utility Company (Pepco), Hospitals, State Agencies that provide Unemployment benefits, etc.

Part 2: Data Ingestion

Government Agencies and private companies can immensely help us solve Homelessness. By agreeing to share vital information with our System Sunryse they will help a person from not becoming Homeless. Sunryse provides offers various means to push the data securely from the Source to Sunryse database. For example, Utility companies can agree to share Information about a person who missed a payment. Sunryse provides an easy way to achieve this. Utility companies can either install a bookmarklet, fill out a form or make a simple change to their system by integrating a small Javascript code.

Part 3: Detection, Prediction and Prevention

Once the data is entered into our system, we run the data against our regression model to identify medium to high-risk candidates. Organizations like St. Patrick can then take a quick look at the preliminary report and decide to continue forward. By doing that they also believe that the case may have some weight. Sunryse will help them reach out to the person via email or phone requesting for additional information. Sunryse offers various (customizable) inputs that our model may require to accurately predict the likelihood of a person to become Homeless. Once the information is received, Sunryse automatically runs the data and come out with detail report and corresponding suggestions and plan of action.