2.5 Million Citizens. One Intelligent System. AI-Powered Grievance System for Pune Municipal Corporation

AI-powered Citizen Grievance Redressal System that automates complaint classification and routes grievances intelligently across 37 departments for Pune Municipal Corporation.

IndustryeGovernance • Public Services • Smart City
Timeline16 Weeks • 7 Phases • Go-Live Approved
TechnologiesXLM-RoBERTa • NLP • GIS / PostGIS • Agentic AI • OCR • REST API
Scale15 Wards • 37 Departments • 170+ Prabhags • 10,000+ Concurrent Users
The Challenge

A City of 2.5 Million Citizens — and a Complaint System Built for a Different Era

Pune Municipal Corporation manages civic services for one of India's fastest-growing cities. But the system through which citizens reported problems had not kept pace. Every complaint, a broken street light, a water supply failure, or an overflowing drain went through a slow, manual chain that frustrated citizens, overwhelmed officers, and left grievances unresolved for weeks. Every step was manual. Every handoff was a delay. With no automated routing, no SLA enforcement, and no real-time visibility for citizens, complaints fell through the cracks, officers were overwhelmed, and trust in digital civic services was eroding.
Officer Paperwork
7–15Days Average Resolution Time
~60%SLA Compliance Rate
100%Manual Complaint Routing
Manual Complaint RoutingManual Complaint Routing
No SLA EnforcementNo SLA Enforcement
Zero Citizen VisibilityZero Citizen Visibility
No Officer DashboardNo Officer Dashboard
Language BarriersLanguage Barriers
No Data for Decision-MakingNo Data for Decision-Making
The Solution

Integrated Citizen Grievance Redressal System (CMS)

Stark Digital built the Integrated Citizen Grievance Redressal System, a full-stack, AI-powered platform that transforms how Pune's citizens report problems and how PMC responds to them. The system handles the entire complaint lifecycle: from multi-channel intake and AI-powered classification to intelligent routing, SLA enforcement, officer resolution, and real-time citizen updates. At its core is a bilingual AI classification engine, trained on 7,800 real PMC complaint records, that automatically suggests the correct department category for every complaint submitted in English or Marathi. Officers confirm the suggestion or correct it, and the system learns from every correction. No full retraining required.

1

Citizen Files Complaint

Complaint submitted via web, mobile, WhatsApp, call center, or walk-in, in English or Marathi. 24/7 availability across all channels.

2

AI Classifies Instantly

XLM-RoBERTa transformer + SGD adaptive classifier analyzes the complaint text and suggests one of 78 PMC categories with a confidence score.

3

Citizen Confirms or Edits

If confidence is high (>80%), the suggestion is shown for confirmation. If low, the citizen or officer selects the correct category. The system learns from every correction.

4

GIS Routes to Officer

GPS coordinates detect the correct Ward, Zone, and Prabhag via PostGIS. Complaint is automatically assigned to the responsible L1 officer within 30 seconds.

Key Capabilities

Multi-Channel Citizen Portal

Web, mobile, call center, walk-in, and WhatsApp intake — all feeding into one unified system. Bilingual Marathi and English throughout. 24/7 availability.

AI-Powered Auto Routing

NLP-based category suggestion with 90.6% accuracy on 7,800 training records. Complaint mapped to correct department automatically. 85%+ complaints require zero manual routing.

GIS Location Intelligence

GPS + geocoding detects Ward, Zone, and Prabhag automatically. PostGIS spatial engine handles address fallback when GPS is unavailable. Zero manual ward assignment.

SLA Enforcement Engine

Each complaint category has a defined SLA (2–7 days based on type). Automated escalation to L2, HOD, and Commissioner on breach. 60% to 90%+ SLA compliance target.

Citizen Transparency Loop

Real-time SMS and WhatsApp updates at every stage — filed, assigned, in progress, resolved. Citizens can reopen a complaint if unsatisfied with the resolution.

Commissioner Analytics Dashboard

SLA heatmaps, ward performance comparisons, department-level resolution rates, and escalation tracking. Real data for real governance decisions.

The Impact

What Changes When a City Gets Intelligent Infrastructure

For Citizens:

  • Complaint filed in minutes from web, mobile, or WhatsApp
  • AI suggests the right category automatically no guessing
  • Real-time SMS and WhatsApp updates at every stage
  • Complaint resolved in 2–5 days not 7–15
  • Transparent status tracking, no more calling the office
  • Reopen mechanism if resolution is unsatisfactory

For Officers:

  • Structured complaint inbox, no more informal assignments
  • Clear SLA deadlines for every complaint
  • AI pre-classifies complaints; the officer just verifies
  • Field evidence upload directly from mobile
  • Automatic escalation paths clear accountability
  • No complaints fall through the cracks

For PMC Leadership:

  • SLA compliance: 60% → 90%+ target
  • 85%+ complaints auto-routed zero manual reception overhead
  • Commissioner dashboard with real-time ward and department analytics
  • Permanent, structured audit trail for every complaint
  • 2.5 million legacy records migrated and GIS-enriched
  • Zero development cost fully funded as CSR by Stark Digital

Your City Deserves This Too

The Integrated Citizen Grievance Redressal System is a replicable model, adaptable for any municipal corporation, state government, or public services organization. Talk to our team about deploying it for your citizens.

DocVerify Visual