Geospatial Data & Search Workflow
Apartment Shopping
Live Web App
A rental shortlist for comparing lease fit, distance, and risk flags while tracking contacts, tours, and decisions.

Key DecisionKeep personal notes intact when the listing catalog changes.
The Case Study
- Problem
- During my search for a short-term Bay Area lease, listings were spread across Zillow, Craigslist, Apartments.com, and building websites. Tabs and spreadsheets made it difficult to keep photos, distances, and the history of contacts, tours, and rejected options together.
- What I Built
- A rental shortlist with lease fit, availability, market comparisons, risk flags, search status, and side-by-side comparison. Offline location lookup makes distance comparisons possible without a geocoding request for every listing.
- How It Works
- A separate engine handles ranking, location lookup, filters, and export. Its clock can be controlled in tests. Stable listing identifiers connect the catalog to notes stored in the browser, so catalog updates preserve the user’s notes.
- Where It Stands
- Built for my own rental search and publicly available as a single-user app. I adapted an existing comparison workflow and its data model to the needs of a rental search.
- Evidence and Limits
- Distance is straight-line, not driving or transit time. Market comparisons are estimates.
- Listings are curated with assistant support. Scam and income-restriction flags are recorded at capture time, not detected by an in-app AI service.
- The test report and review findings provide dated evidence; engine tests do not establish every UI behavior.
- Product Stack
- React · TypeScript · Vite · Vitest · Geospatial (haversine) · GitHub Actions
- Evidence As Of
- 2026-10-27