Hackathon Winner Project
Urban Intel AI
Hybrid AI smart city governance platform with local LLM policy engine
Project Overview
Urban Intel AI is a Smart City Governance Platform I built for the Ingenius Hackathon 7.0 at Ahmedabad University, where it proudly won 1st Runner Up among 180+ participating teams.
The core of the system uses a Hybrid AI Architecture consisting of 6 specialized Random Forest models trained to predict critical urban risks—including water scarcity, traffic congestion, and health hazards—with extremely high precision.
Platform Highlights
// 01
Hybrid AI Architecture
Utilizes 6 specialized Random Forest models trained to predict critical urban risks including water scarcity, traffic congestion, and health hazards with high precision.
// 02
Local LLM Policy Engine
Integrates a private local TinyLlama LLM that processes risk outputs and generates real-time, actionable policy recommendations for city administrators.
// 03
100% Data Sovereignty
Operates entirely without relying on external APIs, ensuring both incredible speed and total privacy for sensitive government and administrative data.
// 04
Hackathon Recognition
Secured 1st Runner Up among 180+ competing teams at the Ingenius Hackathon 7.0 held at Ahmedabad University.
The Unique Edge
What makes Urban Intel AI unique is the integration of a private local TinyLlama LLM that processes the risk outputs and generates real-time, actionable policy recommendations for city administrators. It operates entirely off the grid without external APIs, guaranteeing total data sovereignty, ultimate speed, and absolute privacy for sensitive government data.