GramAI

GramAI is an advanced AI-powered spatial intelligence and rural governance platform. By leveraging high-resolution drone orthophotos, satellite imagery, and deep learning segmentation models, GramAI automates rural property mapping, road network delineation, and municipal tax audit reports.

🚀 Core Solution

GramAI processes geospatial inputs (orthophoto tiles and coordinates) to deliver real-time building footprint extraction, road connectivity analysis, and compliance auditing.

Key Features

• AI-Powered Building & Road Segmentation: Runs a deep-learning UNet++ model for high-precision semantic segmentation of structures, roads, and waterbodies.
• Geospatial Analytics Dashboard: Interactive Mapbox-powered visualization displaying building boundaries, compliance violations, tax leakages, and solar energy candidates.
• Dynamic Pipeline Tracking: Terminal-like progress monitor showing live model execution and inference logs.
• LLM-Powered Compliance Reports: Integrates with Llama to compile structural audit and tax leakage compliance reports, ready for one-click PDF export.

🛠️ Project Setup & Run Instructions

1. Backend (FastAPI)

The backend manages GIS geocoding, local user authentication (SQLite3), UNet++ inference, and LLM-report generation.

• Directory: ./backend
• Prerequisites: Python 3.12+ and uv package manager installed.
• Run Command: bash uv run python -m uvicorn app.main:app --reload --host 127.0.0.1 --port 8000

2. Frontend (Next.js)

The frontend provides a state-of-the-art WebGIS dashboard built with Next.js, React, Mapbox, and Tailwind CSS.

• Directory: ./frontendcd
• Prerequisites: Node.js v18+ installed.
• Run Command: bash npm run dev (Default Port: http://localhost:3000)

Built With

• batchfile
• css
• javascript
• jupyter-notebook
• python
• typescript