GreenGovRAG

November 22, 2025

An Open Source AI Assistant for Australian Environmental & Planning Regulations

Helping planners, consultants, and researchers navigate complex laws — faster, smarter, and cloud-agnostic.

Why GreenGovRAG Exists

Environmental and planning regulations in Australia are fragmented across federal, state, and local governments.

Planners, consultants, and ESG analysts often spend weeks digging through PDFs, council websites, and regulatory guidelines just to answer basic questions like:

How do I calculate fugitive emissions from natural gas transmission pipelines?
(LGA: City of Adelaide) What heritage guidelines apply to developments in Adelaide Park Lands?
How do federal and Queensland planning laws interact for development near Great Barrier Reef?
(Topic: Emissions) How do I report Scope 2 emissions using market-based accounting with renewable energy certificates?

GreenGovRAG changes that.

It’s an open-source AI-powered assistant that makes navigating regulations as simple as asking a question — with citations to official sources and location-aware filtering.

What is GreenGovRAG?

GreenGovRAG is an AI Retrieval-Augmented Generation (RAG) system tailored for Australian environmental and planning regulations.

Key Features:

Ask questions in natural language → get relevant answers with citations
Geo-aware search → filter by state, LGA (Local Government Area), or region
Supports ESG reporting → climate, biodiversity, water, emissions queries
Hybrid search → combines BM25 keyword search with vector similarity
Cloud-agnostic deployment → works on AWS, Azure, or local Docker
Multi-LLM support → OpenAI, Anthropic, AWS Bedrock, Azure OpenAI
Open Source → built for contributors, civic-tech, and researchers

Source code: github.com/sdp5/green-gov-rag

Example Use Cases

1. Environmental Impact Assessment (EIA) Pre-screening

User: Environmental consultant, planner, or developer Query: “Do I need an environmental impact assessment to build a solar farm in regional NSW?”

GreenGovRAG Output:

Summarizes relevant sections from NSW planning portal and EPBC Act
Cites official sources with links
Explains exemption criteria and thresholds

2. Native Vegetation Clearing Rules by Region

User: Local council officer or landowner Query: “Can I clear native vegetation near Murray Bridge, SA?”

GreenGovRAG Output:

Retrieves SA Government vegetation clearance policies
Uses LGA-level filtering for local rules
Returns allowed/disallowed activities and buffer zones

3. Zoning Regulations and Permitted Uses

User: Urban planner or real estate developer Query: “What are the zoning restrictions for coastal land in Mornington Peninsula, VIC?”

GreenGovRAG Output:

Retrieves planning scheme overlays from council documents
Explains permitted uses, height limits, environmental constraints

4. Emissions and Energy Standards Compliance

User: Sustainability advisor or industrial developer Query: “Which emissions standards apply to industrial zones in Greater Sydney?”

GreenGovRAG Output:

Points to NSW EPA and federal requirements
Suggests offsets or sustainable alternatives
Links to energy incentive schemes

How It Works (Technical Deep Dive)

GreenGovRAG combines automated ETL pipelines, RAG, and cloud-native infrastructure:

1. Data Ingestion & ETL

Sources: EPBC Act (federal), SA/NSW/VIC legislation, council planning schemes, emissions data (CER/NGER)
Pipeline: Apache Airflow (dev) + GitHub Actions (production)
Processing: PDF parsing, HTML scraping, metadata tagging with LLM
Storage: PostgreSQL with pgvector, cloud storage (S3/Azure Blob)

2. Text Chunking & Embeddings

Chunking: Semantic chunking with 500-1000 token chunks, 100-200 token overlap
Embeddings: sentence-transformers/all-MiniLM-L6-v2 (default, configurable)
Metadata: Preserves jurisdiction, document type, LGA, regulatory hierarchy

3. Vector Store & Hybrid Search

Vector Stores: FAISS (local dev), Qdrant (production)
Hybrid Search: BM25 keyword search + vector similarity
Geospatial Filtering: LGA boundaries (GeoJSON), location NER for Australian places

4. RAG Chain & Response Generation

LLM Providers: OpenAI (gpt-4o, gpt-4o-mini), Anthropic (Claude), AWS Bedrock, Azure OpenAI
Response Enhancement: Trust scoring, citation verification, regulatory hierarchy
API: FastAPI with OpenAPI docs, rate limiting, CORS support

5. Deployment Options

AWS: ECS Fargate (backend), EC2 Spot (Qdrant), RDS PostgreSQL, CloudFront (frontend)
Azure: Container Apps, PostgreSQL Flexible Server, Blob Storage
Local: Docker Compose with all services
CI/CD: GitHub Actions for automated deployments

Tech Stack:

Backend: Python 3.12, FastAPI, SQLModel, LangChain
Database: PostgreSQL with pgvector
Vector Stores: FAISS, Qdrant
Frontend: React + TypeScript (WIP)
Deployment: Docker, AWS CDK, Azure Bicep

Why It Matters

For Planners

Saves weeks of manual research — get answers to compliance questions in seconds instead of days.

For Consultants

Supports ESG and compliance reporting — quickly find relevant regulations for environmental impact statements.

For Local Councils

Improves transparency and community engagement — make regulations more accessible to residents.

For Researchers

Builds open datasets — contribute to civic-tech and evidence-based policy.

In Adelaide and across South Australia, there’s growing interest from urban planning and sustainability groups who struggle with fragmented regulatory data.

Project Status & Roadmap

Current (v0.1.0):

Multi-source document ingestion (federal, state, local)
Hybrid BM25 + vector search with LGA filtering
Multi-LLM support (OpenAI, Anthropic, Bedrock, Azure)
Production deployments on AWS and Azure
Automated ETL pipelines (GitHub Actions)
API with comprehensive docs

Next Steps:

Geographic coverage expansion — all states and territories
Frontend improvements — interactive query interface
ESG-specific features — carbon accounting, biodiversity metrics
User authentication — OAuth2, usage tracking
Real-time updates — webhook-based document refresh

Success Metrics:

Community adoption → GitHub stars, contributors, PRs
Utility → queries answered correctly, user feedback
Coverage → document sources across all jurisdictions
Engagement → demo usage, conference presentations

How You Can Help

GreenGovRAG is open-source and community-driven. Ways to contribute:

For Developers

Contribute code, tests, or documentation
Add new document source plugins
Improve the frontend UI/UX
Write integration tests

For Domain Experts

Add your state’s/council’s regulations to the ETL pipeline
Validate query results and provide feedback
Suggest new use cases and features

For Advocates

Share with planners, ESG analysts, and researchers
Present at meetups or conferences
Provide feedback via GitHub issues

Documentation: docs.greengovrag.sundeep.id.au
Discussions: github.com/sdp5/green-gov-rag/discussions

Quick Start

Using Docker (Recommended)

git clone https://github.com/sdp5/green-gov-rag.git
cd green-gov-rag/deploy/docker
cp .env.example .env
# Edit .env with your API keys
docker-compose up

Access:

Backend API: http://localhost:8000/docs
Frontend: http://localhost:3000 (WIP)

Local Development

# Backend
cd backend
pip install -e .[dev]
cp .env.example .env
alembic upgrade head
uvicorn green_gov_rag.api.main:app --reload

# Frontend
cd frontend
npm install
npm run dev

See deployment guide for AWS/Azure deployment.

What’s Next?

Expanding coverage across all Australian states and territories
ESG-specific reporting features for carbon, biodiversity, water
Partnerships with civic-tech and green-tech organizations
User authentication and multi-tenancy support

Final Word

GreenGovRAG is our step toward making environmental regulation accessible, transparent, and actionable — starting in Adelaide and scaling across Australia.

Whether you’re a planner navigating complex compliance requirements, a consultant preparing environmental assessments, or a researcher analyzing policy gaps — GreenGovRAG can help.

Join us in building it. 🌏

Generative AI with AWS Bedrock Data + AI