Where developers grow together

ShehrAI

CivicPulse AI What is CivicPulse AI? CivicPulse AI is an AI-powered smart city platform designed to help governments predict infrastructure problems before they become major issues. Instead of waiting for citizens to report potholes, flooding, or damaged roads, the platform analyzes historical and real-time data to identify high-risk locations and recommend preventive actions. The platform combines an interactive Digital Twin map, machine learning, geospatial analytics, and an AI assistant to provide a single system that benefits both citizens and government departments. Why did we build CivicPulse AI? Most cities today operate on a reactive maintenance model. Roads are repaired only after they have already deteriorated and citizens begin reporting problems. This approach leads to delayed repairs, increased maintenance costs, traffic congestion, accidents, and poor use of government resources. We built CivicPulse AI to shift this process from reactive maintenance to predictive maintenance. Our goal is to help governments identify infrastructure problems before they occur, prioritize repairs based on risk, and allocate resources more effectively. At the same time, we want to give citizens better visibility into road conditions and provide them with reliable information for safer travel. How is it useful? For Citizens View an interactive map showing the health and predicted risk of roads. Check road conditions before traveling. Ask an AI assistant questions about road safety, flooding, and travel routes. Report civic issues that are automatically routed to the correct government department. For Government Departments Identify high-risk roads before failures occur. Prioritize maintenance using AI-generated risk scores. Monitor infrastructure through dashboards and heatmaps. Allocate maintenance teams more efficiently. Reduce emergency repairs and improve public safety. How does it work? CivicPulse AI collects infrastructure, weather, and complaint data from multiple sources. The data is cleaned, processed, and stored in Elasticsearch, which serves as the platform's central search and analytics engine. A machine learning model analyzes factors such as rainfall, complaint history, road age, and maintenance records to predict the likelihood of infrastructure failures. These predictions are displayed on an interactive Digital Twin map and are continuously updated as new data becomes available. An AI-powered assistant uses Retrieval-Augmented Generation (RAG) to answer user questions using the latest infrastructure data, ensuring responses are based on verified information rather than generated assumptions. Key Features AI-based prediction of pothole and flood risks Interactive Digital Twin map of city infrastructure Government command center with analytics dashboards AI assistant for citizens using Retrieval-Augmented Generation (RAG) Department-specific complaint routing Real-time geospatial search and visualization Continuous prediction updates as new weather and complaint data arrive Impact CivicPulse AI enables cities to move from fixing problems after they occur to preventing them before they happen. By combining artificial intelligence, geospatial technology, and real-time analytics, the platform helps governments improve maintenance planning, reduce operational costs, enhance public safety, and deliver better civic services while giving citizens greater confidence in their daily travel.

Metro Flow

Metro Flow is a project that was mainly focused on  Smart Coach postioning, Live Crowd Pulse, Last-Mile Transport Stack, Lite AR Wayfinding. Smart Coach positioning defines which coach is near to satirs, Live Crowd Pulse defines which station is busier, Last Mile Transport Stack defines direct booking to home with metro, Lite AR Wayfinding defines a route is displayed on the screen.  

MetroMate AI – Intelligent AI Commute Assistant for Delhi Metro

MetroMate AI is an AI-powered intelligent commute assistant that transforms Delhi Metro travel through personalized route planning, fast station search, and real-time metro health insights. The platform combines official GTFS metro data, graph-based route optimization using Dijkstra's algorithm, Elasticsearch-powered autocomplete with typo tolerance, and Google's Gemini AI to help commuters choose the best route based on travel time, cost, crowd levels, safety, and convenience. Users can ask natural language questions like "Get me to Rajiv Chowk before 9 AM under ₹70" or "Is the Yellow Line delayed right now?" The system retrieves the best routes, evaluates them using the MetroMate Score, provides live operational insights through a simulated health engine, and explains recommendations in plain English. Built with Next.js, Express, TypeScript, Elasticsearch, OpenStreetMap, and Gemini, MetroMate AI demonstrates how AI and