The Problem
Traditional urban public transit in tier-2 cities like Visakhapatnam suffers from a lack of real-time visibility, unpredictable wait times, manual ticket auditing, and frequent commuter grievances (such as skipped stops or overcrowding) with no verifiable proof.

The Solution
VizagBus AI is an intelligent, automated transit telemetry and grievance auditing platform that bridges physical transport operations with digital intelligence.

How It Works (End-to-End Workflow)
Real-Time Data Ingestion:

The platform processes live operational telemetry through a serverless SQLite database backend and FastAPI endpoints.

It ingests transactional data directly from ticketing machines or operational updates, instantly calculating seat occupancy and crowd percentages.

Interactive Commuter Interface:

Commuters access a bilingual web portal (English and Telugu) via any browser.

Using spatial distance calculations (Haversine formula) and Leaflet map rendering, the platform matches user locations to nearby stops, plots walking paths, and displays real-time transit schedules.

AI-Driven Transit Assistance (Gemma 4):

Commuters can query the built-in AI assistant to get instant, context-aware routing advice.

Rather than relying on static responses, the system performs dynamic in-context data injection—passing the live state of the transport network directly into Gemma 4 to recommend optimal travel options and warn against overcrowded routes.

Autonomous Grievance Auditing:

When a commuter submits a complaint (e.g., alleging a stop was skipped despite available space) along with optional media evidence, the backend automatically captures a snapshot of the operational logs at that exact timestamp.

Gemma 4 cross-references the passenger’s claim against the verified telemetry logs, determines validity, calculates regulatory fines, and generates structured disciplinary orders for depot managers.