Citizen Voice AI - Government Accountability System AI Agent That Never Lets Government Ignore You

Project ID: T3:315-Citizen Voice AI MultiAgent Government Accountability System-vinay-vinayb1021

Complete AI-Powered Government Complaint Processing System with Dual Deployment Modes

---

Project Overview

A comprehensive AI system that automates government complaint processing from citizen submission to resolution tracking. Features 6 intelligent agents working together to ensure transparency, accountability, and efficient resolution of citizen grievances.

6-Agent Workflow:

Citizen Complaint → Chat Agent → Router Agent → Tracker Agent
                                                    ├→ Follow Agent
                                                    ├→ Analytics Agent  
                                                    └→ Escalate Agent

Status Progression:

• 🔴 RED: New complaint received
• 🟠 ORANGE: Routed to department with deadline
• 🔵 BLUE: Acknowledged by department
• 🟢 GREEN: Under resolution/progress
• ⚫ BLACK: Resolved and closed

---

Dual Deployment Architecture

Mode 1: Standalone FastAPI Demo

Perfect for: Development, demos, local testing
Features: Instant setup, no API keys needed, full web interface
Deployment: Single command - python main.py

Mode 2: IBM Watsonx Orchestrate Production

Perfect for: Government deployment, enterprise scale
Features: Professional AI orchestration, cloud scalability
Deployment: Import ready-made agents, flows, and tools

---

Prerequisites

• Python 3.9+
• Git (for cloning repository)
• Web Browser (Chrome/Firefox recommended)
• IBM Watsonx Account (for Mode 2 only)

---

Installation Steps

1. Clone Repository

git clone https://github.com/vinay1359/citizen-voice-ai.git
cd citizen-voice-ai

2. Create Virtual Environment

# Windows
python -m venv venv
venv\Scripts\activate

# Mac/Linux  
python3 -m venv venv
source venv/bin/activate

3. Install Dependencies

pip install -r requirements.txt

4. Environment Setup (Optional for Mode 1)

# Copy environment template
copy ".env sample" .env

# Edit .env file with your credentials (only needed for Mode 2)

---

Build & Deployment

Mode 1: FastAPI Standalone Demo

Quick Start (30 seconds):

python main.py

Access Points:

• Citizen Interface: http://localhost:8000
• Government Dashboard: http://localhost:8000/dashboard.html
• API Documentation: http://localhost:8000/docs

Features Available:

• Real-time complaint submission
• 6-agent processing simulation
• WebSocket live updates
• Department routing and tracking
• Status progression visualization
• Government dashboard for officials

---

Mode 2: IBM Watsonx Orchestrate Production

Prerequisites:

• IBM Watsonx Orchestrate account
• CLI tool installed: npm install -g @ibm/watsonx-orchestrate-cli

Deployment Steps:

1. Import Tools:

orchestrate tools import -k python -f "tools/language_detector.py"
orchestrate tools import -k python -f "tools/text_classifier.py"
orchestrate tools import -k python -f "tools/urgency_analyzer.py"
orchestrate tools import -k python -f "tools/department_mapper.py"
orchestrate tools import -k python -f "tools/deadline_calculator.py"
orchestrate tools import -k python -f "tools/status_monitor.py"
orchestrate tools import -k python -f "tools/reminder_scheduler.py"
orchestrate tools import -k python -f "tools/data_analyzer.py"
orchestrate tools import -k python -f "tools/deadline_tracker.py"

2. Import Knowledge Base:

orchestrate knowledge import -f "knowledge/knowledge_base_config.yaml"

3. Import Agents:

orchestrate agents import -f "agents/chat-agent.yaml"
orchestrate agents import -f "agents/router-agent.yaml"
orchestrate agents import -f "agents/tracker-agent.yaml"
orchestrate agents import -f "agents/follow-agent.yaml"
orchestrate agents import -f "agents/analytics-agent.yaml"
orchestrate agents import -f "agents/escalate-agent.yaml"

4. Deploy Agents:

orchestrate agents deploy --name "Chat_Agent"
orchestrate agents deploy --name "Router_Agent"
orchestrate agents deploy --name "Tracker_Agent"
orchestrate agents deploy --name "Follow_Agent"
orchestrate agents deploy --name "Analytics_Agent"
orchestrate agents deploy --name "Escalate_Agent"

5. Import Workflow:

orchestrate flows import -f "flows/complaint-processing-flow.py"

---

Demo Instructions

Mode 1 Demo Walkthrough

1. Start the Application:

   python main.py

   Wait for: Server started on http://localhost:8000

2. Citizen Interface Demo:

   • Open: http://localhost:8000
   • Submit a complaint: "There has been no electricity in my area for 2 days"
   • Expected Output:
     • Complaint ID generated (e.g., abc123def)
     • Status changes: RED → ORANGE
     • Agent activities displayed in real-time
     • Department assignment shown

3. Government Dashboard Demo:

   • Open: http://localhost:8000/dashboard.html
   • View complaint in department queue
   • Change status from ORANGE → BLUE → GREEN → BLACK
   • Expected Output:
     • Real-time status updates
     • Deadline tracking
     • Department workload display

4. Agent Workflow Observation:

   • Chat Agent: Processes complaint → categorizes as "electricity"
   • Router Agent: Routes to "Delhi Electricity Board" → sets 48hr deadline
   • Tracker Agent: Monitors status → sets up tracking
   • Follow Agent: Schedules reminders
   • Analytics Agent: Generates insights
   • Escalate Agent: Monitors for delays

Mode 2 Demo Walkthrough

1. Access IBM Watsonx Orchestrate Interface
2. Start Chat with Chat_Agent:

   {
     "complaint_text": "Water pressure is very low in our building",
     "citizen_name": "John Doe",
     "phone": "+91-9876543210",
     "location": "Mumbai, Maharashtra"
   }

3. Expected Agent Flow:
   • Chat_Agent → processes and returns complaint_id
   • Router_Agent → routes to Mumbai Water Board
   • Tracker_Agent → sets up monitoring
   • Follow/Analytics/Escalate Agents → work in background

---

Project Structure

citizen-voice-ai/
├── 📄 main.py                 # FastAPI application (Mode 1)
├── 🌐 index.html             # Citizen interface
├── 🏛️ dashboard.html         # Government dashboard
├── 📋 requirements.txt       # Python dependencies
├── ⚙️ .env                   # Configuration file
├── 🤖 agents/                # IBM Watsonx agent configs (Mode 2)
│   ├── chat-agent.yaml
│   ├── router-agent.yaml
│   ├── tracker-agent.yaml
│   ├── follow-agent.yaml
│   ├── analytics-agent.yaml
│   └── escalate-agent.yaml
├── 🔄 flows/                 # Workflow definitions (Mode 2)
│   ├── complaint-processing-flow.py
│   └── complaint-processing-flow.yaml
├── 🛠️ tools/                 # Processing tools (Mode 2)
│   ├── language_detector.py
│   ├── text_classifier.py
│   ├── urgency_analyzer.py
│   └── [6 more tools]
└── 📚 knowledge/             # Knowledge base (Mode 2)
    ├── complaint-categories.json
    ├── department-contacts.json
    └── [configuration files]

---

Configuration

Environment Variables (.env)

# IBM Watsonx Configuration (Mode 2 only)
WATSON_JWT_TOKEN=your_jwt_token_here
WATSON_INSTANCE_ID=your_instance_id
WATSON_REGION_CODE=us-south
WATSON_MODE=auto

---

Key Features

Citizen Experience

• Multilingual complaint submission (English, Hindi, regional languages)
• Real-time status tracking with visual indicators
• SMS/Email notifications for updates
• Community visibility for public complaints
• Upvoting system for community issues

Government Experience

• Centralized complaint dashboard
• Department-wise workload distribution
• Automated deadline management
• Escalation alerts and workflows
• Performance analytics and insights

AI Capabilities

• Automatic complaint categorization
• Department routing based on complaint type
• Urgency assessment and prioritization
• Deadline calculation with government SLAs
• Pattern recognition for systemic issues
• Automated escalation for delays

---

Expected Demo Outputs

Successful Complaint Processing:

✅ Complaint ID: GVT2024001
✅ Category: Electricity
✅ Urgency: HIGH  
✅ Department: Delhi Electricity Regulatory Commission
✅ Acknowledgment Deadline: 24 hours
✅ Resolution Deadline: 5 days
✅ Status: ORANGE → Routed to Department
✅ Tracking: ACTIVE
✅ Reminders: SCHEDULED

Real-time Updates:

• Agent status indicators in UI
• WebSocket notifications
• Status progression visualization
• Department workload updates
• Escalation alerts when needed

---

Deployment Options

Local Development

python main.py  # Runs on localhost:8000

Troubleshooting

Common Issues:

"Module not found" error:

pip install -r requirements.txt

Port 8000 in use:

# Change port in main.py or kill existing process
netstat -ano | findstr :8000

WebSocket connection failed:

• Refresh browser
• Check firewall settings
• Ensure main.py is running

---

License

MIT License - Open for educational and government use

---