Sankalpa - AI-Powered Personalization for SBI Life Insurance

Problem Statement

Enhancing Customer Experience and Retention through AI-Driven Personalization to Improve the Propensity to Purchase

---

🧠 Overview

Developed an AI Driven Personalization Platform for SBI Life as part of a hackathon project. The goal was to enhance policy personalization, upselling, customer engagement, and trust through AI-powered solutions. The platform combines recommendation engines, AI chatbots, gamification, real-time analytics, and ethical AI practices to create an interactive and transparent customer experience.

---

🖼 Screenshots of Interface

---

✨ Key Features

🔐 Authentication & Security

• Dual Login System: Separate portals for customers and agents
• Multi-Factor Authentication (MFA): TOTP-based security using Google Authenticator
• Profile Management: Comprehensive user profile setup for first-time users

🎯 Smart Policy Recommender

• RAG-based Policy Recommendations: Personalized insurance policy suggestions using Retrieval-Augmented Generation
• Trust Policy Score: Explainable AI that builds customer confidence through transparent recommendation reasoning
• Policy Upselling Engine: AI-powered cross-selling and upgrade suggestions to maximize customer value

🤖 TalkSmart Chatbot (TIA)

• Multilingual Support: Contextual customer support in multiple languages
• Intelligent Query Handling: Advanced NLP for understanding complex customer queries
• 24/7 Availability: Round-the-clock customer assistance

🎮 Interactive Financial Tools

• Financial Simulator: A gamified tool for financial planning and scenario modeling that helps users explore different financial strategies in an engaging way.
• Goal GPS: An interactive goal-setting and tracking game that allows users to set, adjust, and achieve their financial objectives through a fun experience.
• AI Powered Nudges: Intelligent, real-time suggestions during gameplay to help users adjust goals, amounts, and timelines for improved financial outcomes.

📊 InsureInsights Dashboard

• Predictive Analytics: Advanced insights for business intelligence
• Campaign AI Optimization: Automated marketing campaign enhancement
• Real-time Engagement Metrics: Live tracking of user interactions and performance
• Inspirational Stories: Customer success stories to drive motivation

💬 FinChat Assistant

• General Financial Knowledge: Comprehensive financial information chatbot
• Educational Content: Answers to financial queries and learning resources
• Contextual Responses: Tailored information based on user profile and history

---

🏗 Model Architecture

---

📁 Project Structure

├── public/ # Static files
├── views/ # Frontend views
├── app.js # Main backend server
├── chatbot_server.py # Chatbot interaction logic
├── dashboard.py # Backend dashboard and visualizations
├── dbSetup.js # Database initialization
├── policy_classifier.py # ML model for policy prediction
├── policy_recommend.py # Core AI logic for recommending policies
├── start_chatbot.js # Starts the chatbot service
├── start_streamlit.js # Launches Streamlit frontend (optional)
├── goal_gps.js # Financial GPS module
├── totpUtils.js # MFA Support with TOTP
├── trust_policy.py # Ethics and fairness enforcement
├── upsell_predictor.py # Predicts upselling opportunities
├── sbilife.xlsx # Sample financial data
└── README.md

---

🔄 User Flow

Customer Journey:

1. Login/Registration → Profile Setup (first-time users)
2. Dashboard Access → Feature Selection
3. Smart Recommendations → Policy Exploration
4. Interactive Tools → Financial Planning
5. Chat Support → Query Resolution
6. Insights & Analytics → Progress Tracking

---

📈 Business Impact

• Increased Customer Engagement: Gamified tools drive user interaction
• Improved Conversion Rates: AI-powered recommendations boost sales
• Enhanced Customer Satisfaction: Multilingual support and personalization
• Data-Driven Decision Making: Comprehensive analytics and insights
• Operational Efficiency: Automated processes and intelligent routing

---

⚙️ Tech Stack

• Backend: Node.js, Python, REST API
• Frontend: HTML, CSS, JS, Streamlit
• AI/ML: Scikit-learn, Sentiment Analysis, RAG
• Database & Security: SQLite, TOTP (Time-based One-Time Password)

---

🎯 Future Scope

Phase 1: Enhanced AI Capabilities

• Advanced ML Models: Implementation of deep learning for better predictions
• Voice Integration: Voice-enabled chatbot interactions
• Computer Vision: Document analysis and claim processing automation
• Blockchain Integration: Secure policy management and claim verification

Phase 2: Platform Expansion

• Mobile Application: Native iOS and Android apps
• API Marketplace: Third-party integrations and partnerships
• Microservices Architecture: Scalable system redesign
• Cloud Migration: AWS/Azure deployment for better scalability

Phase 3: Advanced Features

• IoT Integration: Smart device data for personalized pricing
• Augmented Reality: Virtual policy consultations
• Predictive Analytics: Advanced risk assessment models
• Social Features: Community-driven financial planning

Phase 4: Global Expansion

• Multi-Currency Support: International market expansion
• Regulatory Compliance: Region-specific insurance regulations
• Localization: Cultural adaptation for different markets
• Partnership Ecosystem: Insurance provider network expansion

Phase 5: Innovation Lab

• Quantum Computing: Advanced risk calculations
• AI Ethics Framework: Responsible AI implementation
• Sustainability Metrics: Environmental impact tracking
• Research & Development: Continuous innovation pipeline

---

👥 Team

• Paranthagan S
• Pranaav U
• Padmapriya C
• Nandana M