Boumellassa Haythem

working on my final-year project at VERMEG, developing a modular, microservices-based platform for non-life insurance claims management. The platform automates claims processing, optimizes indemnifications, detects fraud, and enhances efficiency using Spring Boot, AI, and applied mathematics. Key Responsibilities 🔹 Microservices Development: Designing and implementing scalable microservices for claims declaration, tracking, indemnification, and fraud detection, ensuring seamless integration and real-time processing. 🔹 Risk Assessment & Indemnity Calculation: Developing mathematical models for risk evaluation, cost prediction, and indemnification based on policy terms and historical claims data. 🔹 Fraud Detection: Applying machine learning techniques like anomaly detection, clustering (K-Means, DBSCAN), and Bayesian models to identify fraudulent claims. 🔹 Optimization & Decision-Making: Utilizing Monte Carlo simulations, Markov decision processes, and linear optimization to refine claims processing, minimize delays, and reduce financial losses. 🔹 Scalable Deployment & Monitoring: Implementing Kubernetes, Kafka, and Elasticsearch for efficient microservices orchestration, event-driven processing, and real-time data indexing. 🔹 Security & Authentication: Integrating Keycloak for secure identity and access management for clients and insurance agents. Tech Stack & Tools ✔ Backend: Spring Boot, Spring Cloud, Hibernate ✔ Databases: PostgreSQL (relational), MongoDB (NoSQL), Elasticsearch (real-time search & analytics) ✔ Machine Learning & Analytics: Python (NumPy, SciPy, Scikit-learn, TensorFlow) for risk assessment, cost prediction, and fraud detection ✔ DevOps & Deployment: Kubernetes, Kafka, Prometheus & Grafana This project leverages AI, big data, and cloud computing to improve fraud detection, indemnification accuracy, and customer satisfaction, shaping the future of insurtech solutions.