2025 · CHAPA · BU Spark! DS701
Affordable housing access
As part of a collaboration with CHAPA and BU Spark!, I worked with classmates to analyze property data and provide insights about applicant demographics. My primary contribution to the project was an interactive map of affordable housing applications across Massachusetts, showing applicant flow lines, demographic breakdowns, and more.
The map
The tool is a Folium and Leaflet application over four toggleable layers: 109 properties, the applicant origins feeding them, and roughly 750 lines connecting the two. Each property is a pie chart placed at its location, sized by how many applications it received and divided by the racial and ethnic composition of its applicant pool. This makes it possible to roughly evaluate property popularity and applicant demographics at a glance.
The layer with the most utility is the flow lines. Every applicant origin is joined to the property they applied to, which helps visualize both the distance and diversity of applicant hometowns to specific properties. Hovering over a flow shows a popup with information on the count from that origin, the age distribution, and the demographic breakdown of the people it represents. Opening a property gives its details alongside its applicant statistics: median age and range, median distance travelled and range, and the demographic composition as proportion bars.
The remaining layers are applicant origins on their own and a price heatmap, and there's a filter panel over the whole thing — minimum applications, maximum price, maximum distance, maximum age — so the map can be narrowed down for easy viewing. Filtering applies across every layer at once, which is where most of the custom JavaScript comes from; Folium will render the layers, but the cross-filtering, the flow highlighting, and the pie markers all had to be written on top of it.
Context
CHAPA is the umbrella nonprofit for affordable housing and community development in Massachusetts, monitoring close to three thousand permanently affordable homes. The team's brief for the semester was geographic trends, demographic representation, and access disparities in the homeownership portfolio: comparing where applicants lived against where they applied, measuring the radius of their search, and testing how those patterns differed across families, seniors, and racial groups. We also looked at whether marketing strategy, listing source, or the complexity of the application itself shaped who applied at all.
The project ran through DS701 with BU Spark!, which pairs teams with external partners for a semester, and was presented at Demo Day in December 2025.