SmartPost AI is an AI-powered grievance management system designed to help large-scale postal services like India Post automatically analyze, categorize, and prioritize customer complaints using rule-based NLP with sentiment analysis.
The system reduces manual effort, improves response time, and ensures that critical issues are addressed first.
Postal departments receive thousands of complaints every day related to:
- Delivery delays
- Lost or damaged parcels
- Staff-related issues
- Urgent or sensitive cases
Manual classification and response handling often result in:
- Delays in resolution
- Poor prioritization
- Operational inefficiencies
Build an AI-powered system that:
- Automatically analyzes complaint text
- Classifies complaints into predefined categories
- Detects urgency and sentiment
- Suggests appropriate automated responses
- Helps administrators prioritize critical cases
SmartPost AI acts as an intelligent middleware layer between citizens and postal authorities. It processes complaint text using rule-based NLP with sentiment analysis to generate structured, actionable insights.
- Categorizes complaints into:
- Lost Parcel
- Delivery Delay
- Damaged Item
- Staff Behavior
- Uses weighted keyword matching and contextual rules.
- Sentiment scoring using AFINN-based logic
- Automatically marks complaints as CRITICAL when negative sentiment combines with urgent keywords (e.g., medicine, passport, emergency)
- Generates category-based, empathetic response templates
- Reduces manual drafting effort for officers
- Basic email notifications (OTP / alerts) are supported; advanced notification workflows are planned as future enhancements.
- Admins can correct misclassified complaints
- System stores corrections to improve future keyword matching logic
- Text-based complaint submission
- Multilingual-friendly input support
- Image upload for damaged parcels
- Complaint tracking interface
- Timeline-style visualization
- Complaint analytics and category distribution
- Smart priority queue (urgency-based sorting)
- SLA breach indicators
- Auto-response suggestions
- Postman / task assignment (prototype logic)
- React 19
- Vite
- Tailwind CSS
- Lucide React
- Node.js
- Express.js
- Sentiment.js (AFINN-165)
- Custom weighted keyword matching
- Rule-based NLP pipeline
- JSON-based local persistence (prototype)
postal-solution/
├── postal-backend/
│ ├── server.js
│ ├── server_v2.js
│ ├── data.json
│ ├── learned_keywords.json
│ └── test scripts
│
├── postal-frontend/
│ ├── dist/
│ ├── assets/
│ └── index.html
│
├── screenshots/
└── README.md
git clone https://github.com/HarshaAppikatla/postal-solution.git
cd postal-solutioncd postal-backend
npm install
node server.jsBackend runs at: http://localhost:5000
cd postal-frontend
npm install
npm run devFrontend runs at: http://localhost:5173
Frontend deployed on GitHub Pages: 👉 https://HarshaAppikatla.github.io/postal-solution
⚠️ Backend APIs run locally for hackathon purposes. Some features operate in simulation mode when the backend is not connected.
- Postal grievance redressal systems
- Government service automation
- Complaint triage platforms
- Hackathon & academic demonstrations
- Workflow prioritization systems
- Persistent database integration (MongoDB / PostgreSQL)
- JWT-based authentication & role management
- Multilingual NLP models
- ML-based classification (TF-IDF / BERT)
- Advanced email & SMS notification workflows
- Admin analytics dashboard
- Cloud deployment (Render / AWS / Vercel)
Developed for POSTA-THON 2025, focused on building intelligent, scalable grievance-redressal systems for public services.
Harsha Vardhan Appikatla Software Developer | Full-Stack & Backend Enthusiast
🔗 GitHub: https://github.com/HarshaAppikatla