Citizens photograph a pothole, garbage, a dark streetlight or a water leak, and Ward Fix does the rest.

The photo is sorted into its kind by an AI model running inside the browser, so it never goes to an AI company. Reports of the same problem within 50 metres merge into one, so the ward office sees one pothole that six people reported, not six potholes. Every kind has a deadline anyone can watch, and the ward office works through a queue that puts the most overdue first.

Built on a real problem: Smart India Hackathon 2025, statement SIH25031 (Government of Jharkhand), crowdsourced civic issue reporting.

It runs on an ordinary laptop with Node.js and nothing else: it starts its own database and fills it with a sample ward, so the demo works the first time. Light and dark themes, and it works on a phone.

Keywords: MERN, civic issues, CLIP, geospatial, SLA, Smart India Hackathon