All systems

Founding engineer, three-engineer team, venture-backed startup. 2025 to 2026.

Geospatial business intelligence platform

A location intelligence platform covering 67.5 million businesses in 30 countries. A site-selection or network-planning team asks what is around a location and gets an answer in seconds, through a dashboard, an API or directly inside an AI assistant.

67.5M
business entities
2.7s to 27ms
spatial query after index work
10M+
pages crawled

How data moves through it

SourcesRefreshCoreAccess27 msBusiness registries30 countriesPlaces and POI data67.5M entitiesWeb enrichmentcontacts, documentsMonthly release diffonly real changesGeo indexhexagonal cellsAI extractiontested against a setPostgreSQL75M rows, live refreshDashboardmaps, lists, chatREST APIkeys, billingMCP serverinside AI assistants

Relevant to you if

You sell or plan against location data, your analysts answer the same questions by hand every week, or you want your data usable from inside AI tools, not only from a dashboard.

What it changed

Two enterprise pilots secured, with a global real-estate developer and a national telecom operator, on a platform a three-engineer team took from empty repository to pilot-ready in two months. The data model, API, MCP server, crawler infrastructure and dashboard were my part.

How it works

  • One database of 67.5 million businesses across 30 countries, indexed by location, so a question about everything within ten minutes of a site is answered in milliseconds instead of by a study.
  • The whole corpus refreshes every month without taking the product offline, and only real changes reach customers, so the data stays current and quiet.
  • A crawler enriched millions of records with contact details and documents, with AI extraction checked against a test set before anything went live.