Where does NYC's 311 system fall behind?
A relational database and Power BI dashboard built on real NYC 311 service request data, with an automated Python and SQL pipeline behind it.
The question
NYC's 311 line handles millions of service requests a year, everything from noise complaints to broken streetlights. Response times vary a lot by agency, borough, and complaint type, but that variation is buried in raw city data. This project turns it into something a decision maker could actually act on.
Replace this paragraph with your specific question once you've explored the data, for example: which agency misses its own response time the most, or which neighborhoods wait longest for the same complaint type.
How it works
Pull
A Python script calls the NYC Open Data (Socrata) API for 311 requests.
Automate
A scheduled GitHub Actions workflow runs that script weekly, no manual step required.
Model
Data lands in a PostgreSQL star schema: one fact table, four dimension tables.
Analyze
SQL with CTEs and window functions answers questions like agency ranking and month over month change.
Visualize
Power BI connects to the database directly and models relationships, DAX measures, and the dashboard below.
Data model
Key finding
Fill this in once you've run the analysis. A specific number beats a general claim, for example a percentage gap between the slowest and fastest agency, or a resolution-time trend over time.
Live dashboard
