br/acc offers an open-source graph infrastructure to ingest, normalize, and query Brazil's fragmented public databases, generating actionable intelligence for civic improvement. It targets researchers and power users by consolidating scattered open data into a single, queryable graph, enabling users to derive their own insights without pre-interpretation.

How It Works

Over 45 ETL modules ingest and normalize data from diverse Brazilian public sources into a Neo4j graph database. A FastAPI backend provides a public API, complemented by a React frontend for exploring corporate networks and entity connections. The system prioritizes privacy, adhering to LGPD compliance with configurable public-safe defaults.

Quick Start & Requirements

Local setup uses Docker Compose. make bootstrap-demo provides a quick start with seed data. Full ingestion via make bootstrap-all is resource-intensive, potentially taking hours and requiring substantial disk, memory, and network bandwidth. API health is at http://localhost:8000/health, frontend at http://localhost:3000.

Highlighted Details

• Features 45+ implemented ETL pipeline modules.
• Utilizes Neo4j graph database and FastAPI backend.
• Includes a React frontend for interactive data exploration.
• Offers one-command local bootstrap for reproducibility.
• Employs privacy-first design compliant with LGPD.

Maintenance & Community

Driven by a "decentralized movement of Brazilian builders," community engagement occurs via Discord and Twitter (@brunoclz). Specific details on core maintainers or sponsorships are not provided.

Licensing & Compatibility

Licensed under GNU Affero General Public License v3.0 (AGPL v3.0). This strong copyleft license mandates that derivative works or network-accessible services must also be AGPL v3.0 licensed, potentially restricting integration into closed-source commercial products.

Limitations & Caveats

The project website is noted as "in progress." Full ingestion is resource-intensive and time-consuming. The repository does not include a pre-populated production Neo4j dump or guarantee external data source stability. Operational runbooks for institutional use are also excluded.