Apppublic

atharavdhumone/terra-mind

sourceHugging Faceupdated 6d agoView on Hugging Face

0likes

App README

TERRA-MIND: AI Biodiversity & Ecological Intelligence Engine

Darukaa.Earth Hackathon Submission

Multi-agent AI system for ecological reasoning built on LangGraph, ChromaDB, and Google Gemini 2.5. Delivers scientifically grounded, citation-backed recommendations for biodiversity restoration.

Architecture

User Input -> LangGraph StateGraph
  triage_node -> [incomplete] -> clarification_node -> END
              -> [complete] -> spatial_reasoning_node -> retrieval_node -> synthesis_auditor_node -> END
-> Pydantic v2 JSON Output (Recommendations + Citations + Metric Deltas + Confidence)

Component	Technology
Agent Orchestration	LangGraph StateGraph
LLM	Gemini 2.5 Flash/Pro (configurable)
Knowledge Base	FAO, IPCC AR6, IPBES, ICRAF, IUCN
Vector Store	ChromaDB + HuggingFace Embeddings
Schema	Pydantic v2
API	FastAPI + Uvicorn
UI	Streamlit + Folium
Tests	Pytest

Setup

bash

pip install -r requirements.txt
cp .env.example .env
# Set GOOGLE_API_KEY in .env
python knowledge_base/ingestion_pipeline.py

Run

bash

# FastAPI
uvicorn src.api.app:app --host 0.0.0.0 --port 8000 --reload

# Streamlit (separate terminal)
streamlit run ui/streamlit_app.py

Test

bash

pytest tests/ -v

Docker

bash

GOOGLE_API_KEY=your_key docker compose up --build

API Example

bash

curl -X POST http://localhost:8000/api/v1/analyze \
  -H "Content-Type: application/json" \
  -d '{"query_text": "Low SOC, low rainfall, monoculture wheat", "soil": {"organic_carbon_pct": 0.3}, "climate": {"annual_rainfall_mm": 480}, "land_use": {"land_cover_type": "Monoculture Wheat"}}'

LLM Configuration

Set LLM_MODEL in .env to switch models:

—gemini-2.5-flash (default, fast)
—gemini-2.5-pro (deeper reasoning)

Author

—Atharav Dhumone