Use PySpark to process telemetry data for risk modelling

In this talk, we will explore how PySpark can be applied to large-scale telemetry data to build predictive risk models. Drawing on the Motor Vehicle Incident (MVI) project, Linzi will walk us through the end-to-end implementation of a Big Data solution on Azure and Databricks. By integrating Geotab telemetry with Wolters Kluwer Enablon data, the project analyzed correlations between driving behavior and motor vehicle events, enabling the prediction of incident likelihood across multiple dimensions. Attendees will gain insight into handling massive datasets, applying distributed processing frameworks, and turning raw data into actionable risk intelligence.

Our speaker
Linzi Jiang is a Staff Data Scientist at one of North America’s largest energy delivery companies, based in Calgary, Canada. With over 14 years of experience spanning data analytics, machine learning, and digital transformation, she has led impactful projects in both the energy and consulting sectors. Beyond her professional work, Linzi is an avid outdoor enthusiast who enjoys hiking, cycling, and skiing, as well as reading and traveling the world.