BlackPrint Technologies · Jun 2026
State Geo-Intelligence Platform
Economic-Census Explorer with a RAG Agent
Overview
Prototyped a state-level geo-intelligence platform for a Mexican state's sustainable-development agency that puts environmental, industrial and public-program modules over a single economic-census map, with roles, an audit trail and an AI agent that cites its sources.
Highlights
- 01
Implemented three operational modules across the state's 18 municipalities: environmental inspections with 16 category filters and offline field capture, industrial-park mapping with SME–industry matching, and a five-program beneficiary and coverage suite.
- 02
Built a RAG agent with source citation and guardrails that answers through Claude when configured and falls back to a local BM25 retriever, on web chat and WhatsApp.
- 03
Designed a data layer that switches between seed data and Supabase (PostGIS/pgvector with row-level security) and server-side filtering that sustained a 220,000-unit synthetic census.
Stack
- Next.js 16
- React 19
- TypeScript
- MapLibre GL
- Supabase
- PostGIS
- pgvector
- Row-Level Security
- Claude API
- BM25 RAG
- PWA (Dexie/IndexedDB)
- WhatsApp Business Webhooks
Working on something similar?
Tell me about the problem and the data behind it. I reply within 24–48 hours.
Discuss a project