Development of an Integrated Alert System for Earthquake Parametric Insurance

This report details the design and implementation of an automated Alert System structured to support the Earthquake team at Descartes, an Insurtech company specializing in parametric insurance against natural disasters. The system was built to provide daily notifications of relevant earthquakes worldwide and automate the claims calculation process, i.e. the process of detecting and computing eventual losses caused by recent events. The project was initiated to address the need for efficient monitoring of global seismic activity, particularly in regions with high insurance exposure, and to streamline the labor-intensive and delay-prone process of claims assessment following an earthquake. The Alert System is structured as an end-to-end data project, encompassing data ingestion, parsing, processing, and visualization. It is designed to run each morning as part of the data engineers’ Airflow DAG, ensuring that daily updates are consistently delivered to the team. The development process followed an iterative approach, beginning with a Proof of Concept that was refined based on feedback from key stakeholders, including underwriters and risk managers, and culminating in the industrialization and integration of the Alert System into the Earthquake team’s codebase. The report discusses the system’s data sources, architecture, validation process, and future improvements, while also providing context about the fields of Para metric Insurance and Seismology where the Alert System operates. Ultimately, it demonstrates the benefits of integrating advanced data processing and automation within the insurance industry.