MK MetaForge

CivicSight is a multimodal civic intelligence platform designed to help understand what is happening across a city, why it is happening, and what might happen next. It brings together different signals such as weather conditions, traffic observations, citizen reports, and visual evidence to build a more connected picture of civic incidents. Residents can report issues such as potholes, waterlogging, road obstructions, and unusual crowding, while the system uses visual analysis to extract structured observations from uploaded images and videos. The platform combines these signals through an event-driven backend, maintains evidence provenance, and presents the resulting information through an interactive city dashboard. It also includes a machine-learning component that estimates the likelihood of a traffic disruption within the next 30 minutes, alongside a structured workflow for tracking citizen reports. A key part of CivicSight is distinguishing between real external data, citizen-reported information, model-inferred evidence, historical observations, and simulated demo signals rather than presenting everything as verified ground truth. The goal is to move beyond isolated alerts and build a system that connects observations, context, and predictions to support better civic awareness and response.