Insurance Management System with AI

Insurance Management System with AI

Insurance Management System with AI is a Django-based web application named SecureLife. The project combines insurance policy management with customer services, premium payments, claims handling, support questions, and a custom Python machine-learning engine.

The application uses Django 5.1.15 for the web framework, SQLite for data storage, and Razorpay for premium payment processing. The AI functionality is implemented directly inside the project without NumPy or scikit-learn.

Key Features

• Customer registration and profile management
• Insurance policy browsing and category filtering
• AI-based policy recommendations
• AI-powered premium estimation
• Customer risk profiling
• Insurance policy applications
• Premium payment processing through Razorpay
• Payment history and printable receipts
• Insurance claim submission and tracking
• AI claim fraud screening
• Aria AI support assistant
• Support question sentiment analysis
• Support question urgency analysis
• AI-drafted replies for administrators
• Customer segmentation using K-Means
• AI Insights dashboard
• Customer activity history
• Help desk question management

AI and Machine Learning Features

The project contains a custom AI engine under the ai application. The Aria Support Assistant uses TF-IDF, Multinomial Naive Bayes, and cosine nearest-pattern matching to answer support questions. It can also use customer policy, payment, due, and claim information when responding to personal questions.

The sentiment and urgency analysis functionality uses Multinomial Naive Bayes with keyword rules to classify support questions and assist with question triage.

The Policy Recommender uses Softmax multinomial logistic regression with SGD to generate personal match scores for insurance plans and support cross-sell suggestions.

The Premium Estimator uses log-linear ridge regression with a closed-form approach to provide premium estimates and policy price checks.

The Risk Profiler uses logistic regression to calculate a health and claim risk score from customer profile information. The Claim Fraud Screen also uses logistic regression to calculate a risk score for submitted claims and provides a risk label and supporting factors.

Customer segmentation is performed using K-Means with k-means++ for the administrator AI Insights page.

Customer Portal

The customer portal provides access to the main insurance services. Customers can maintain their profile, browse and filter policies, apply for insurance policies, make premium payments, view payment history, submit claims, track claim status, use the Aria AI assistant, receive recommendations, estimate premiums, and submit support questions.

The dashboard displays information including total cover, premiums due, AI recommendations, risk score, and recent activity.

Policy and Application Management

Customers can browse policies using search and category filtering. Policies can also be ranked using AI match results. Individual policy pages provide an AI price check, while duplicate policy applications are blocked.

Administrators can manage policy categories and insurance policies, review applications, and approve or reject applications. Separate views are available for approved, disapproved, and waiting policy holders.

Premium Payment

Razorpay is integrated for premium payment processing. The payment flow supports UPI, cards, net banking, and wallets. Payment records contain transaction details such as amount, currency, gateway, receipt number, Razorpay order information, payment information, status, and payment date.

The server creates payment orders and verifies the Razorpay HMAC-SHA256 signature before marking a premium as paid. A demo checkout is available when Razorpay keys are not configured.

Claims Management

Customers can submit supported insurance claims by providing the incident date, claimed amount, and description. Claims can then be tracked through their status.

The system stores an AI risk score, risk label, and related factors for each claim. Administrators can review, approve, reject, or rescore claims from the claims queue.

Admin Console

The administrator console provides tools for managing customers, policy categories, insurance policies, applications, payments, claims, and support questions.

The business dashboard displays revenue information, applications by category, AI alerts, and pending approvals. Administrators can also access detailed customer information containing policy, payment, claim, and cross-sell information.

AI Insights

The AI Insights page provides the model registry and metrics along with triage information, assistant analytics, demand forecasting, and customer segmentation information.

Database

The application uses SQLite through Django's database configuration. Important models include Category, Policy, PolicyRecord, Payment, Claim, Question, Customer, and ChatLog.

Technology Stack

Testing

The project includes tests for the insurance application. The documented testing command is python manage.py test insurance.

Watch Project Demo

Watch the project demonstration to understand the customer portal, insurance management features, AI functionality, payment workflow, claims management, and administrator console.

Watch Insurance Management System with AI Demo

More Details and Full Documentation

For complete project information, technical details, implementation instructions, and documentation, visit the original UpdateGadh project page.

View Complete Insurance Management System with AI Documentation