← Back
CHALLENGE
Mapping Citi Bike's Operations Against the Neighborhoods It Serves
Citi Bike's public GBFS feed updates every 30–60 seconds: how many bikes, how many open docks. But a station showing ten bikes and ten docks tells you nothing about whether those bikes will be gone in twenty minutes, or whether that station is the only mobility option for a neighborhood a quarter-mile from the nearest subway The challenge was building a tool that could surface which stations needed attention, where service gaps fell along equity lines, and how to reason about rebalancing priorities without access to proprietary data
Client
Better Cities Lab Research
Location
New York City, NY
Year
2026
I defined the central inquiry, shaped the feature set, directed the visual identity, and drove the intellectual framing: the equity lens, the pressure scoring methodology, and the expansion zone overlay as policy communication
THEMES
mobility · spatial data analysis · data-informed strategy · civic tech · climate resilience · public life
HIGHLIGHTS

01
A Hypothesis: Put Performance and Context in One Frame
Combine live ridership data with contextual signals namely transit proximity, population density, bike lane access for a pressure model that tells stakeholders why a station is experiencing pressure and which stations have structural causes that won't be fixed by a single rebalancing run

02
Pressure Score
A composite 0–100 index that updates in real time. It weights dock occupancy extremity most heavily, then subway distance, population density, and lane access. The Methodology tab offers every weight as an interactive slider

03
Equity Layer
A choropleth map of NYC census tracts colored by equity need, calculated as a composite of station pressure, transit desert status, and car-free household rate. This surfaces where Citi Bike is most needed by the people least able to substitute for it
I collaborated with Claude Code as my full-stack engineering partner. Translated design and research decisions into working Next.js components, built the live data pipeline and enrichment hooks, implemented spatial indexing for subway proximity, and surfaced technical constraints that shaped the design, including silent failures in the equity model's Census API integration that only appeared during a close audit
Say hello. Let's make magic ↗
← Back

