StreetScope AI

Overview

Geospatial business-intelligence tool built around an Opportunity Finder: type a business category — say, "car wash" — and see where competitors cluster versus the green zones where the category is under-represented, so you can pick a site with less competition. Seeded by a custom Google Places scraper that indexed 215,000+ businesses, heaviest across Florida and scattered across the rest of the US. Paused while I solve the data layer, not killed — the scraper lives on as a public repo and the thesis is still open.

How it works

1. 01Opportunity Finder UI — pick a business type, see competitor density on the map and green zones where the category is thin, so you can pick a site based on where the competition isn't.
2. 02Database seeded by a custom Google Places scraper that indexed 215,000+ businesses — heavy coverage across Florida, sporadic across the rest of the US. State was uneven by design: scraped where I had the budget to scrape.
3. 03Ran straight into the limits of Google Places as a data source. Per-query result caps miss businesses that are literally next door to each other. Places' service-type taxonomy is shallow, so records often came back without the distinction I needed — a car wash and a hair salon can end up looking the same in the database.
4. 04Fully backfilling to the density and category fidelity the Opportunity Finder needs would have meant blowing through both the API rate limits and the per-request cost. Without trustworthy coverage and categories, the green zones were too noisy to hand a customer.
5. 05Paused rather than killed: the product can come back if the data layer gets solved — a cheaper or denser source, a better category taxonomy, or a tighter vertical slice where coverage is already clean.
6. 06Scraper released publicly on GitHub — useful on its own as a Google Maps data tool even without the BI layer on top.

Links