Pop-Up Optimizer
Rank civic-service stops on a Dallas map using need-weighted clustering.
- Role
- Solo, AI-assisted prototype session
- Stack
Problem
Pop-up health vans, food banks and vaccine clinics need locations that balance population, need and travel distance. A transparent placement heuristic makes those tradeoffs inspectable.
What I built
An AI-assisted offline web app offers three services, one to ten stops and adjustable travel windows. Weighted k-means proposes centers, snaps them to sample locations and ranks them by incremental need-weighted coverage.
An SVG geographic map and plain-English explanations show coverage and high-need counts. Shared residents are counted only once in the overall served total.
Results
- The default five-stop health demo covers 96,805 of 269,985 synthetic residents. Population, need indicators and clinic flags are invented demo data, not Dallas demographic estimates.
- Nine optimizer configurations and five HTTP routes pass checks. Real Chromium checks also cover all service controls, map links, a 390px mobile viewport and standalone offline rendering without network requests. The schematic map, approximate travel times and unverified venues limit this to an illustrative demo.