Matthew Erxleben
I'm a data-driven engineer and AI enthusiast, currently completing a BASc in Management Engineering with an AI emphasis at the University of Waterloo.
About
I’m a data-driven engineer and AI enthusiast, currently completing a BASc in Management Engineering with an AI emphasis at the University of Waterloo. I combine engineering rigor with hands-on data work to build scalable dashboards, ML models, and reliable data pipelines.
I’ve applied these skills across Tesla, Electronic Arts, AMD, and academic labs to improve maintenance, user engagement, and wildfire risk prediction. I enjoy turning complex data into clear impact and collaborating across cross-functional teams.
- 4 years' experience
- Joined September 2025
Industry experience
- Software & Internet
- Professional Services
- Media & Entertainment
Work experience· 6 roles
Developed fault-detection queries in SQL on Megapack ECU telemetry, triggering real-time alerts for service technicians; Configured Helm charts to route ECU signals into SQL alert logic, enabling fleet-wide alerting updates; Engineered autoencoder models in Python to uncover anomalous behaviour in telemetry data, enabling preventative maintenance
Processed BC weather/topography data using Python and Pandas to predict wildfire risk; Conducted literature review on Deep Learning for wildfire prediction and co-authored a review paper in ISPRS Journal; Continued research with Waterloo’s AI design team WAT.ai, utilizing ERA5 and MODIS datasets, and testing deep learning models
Developed a player segmentation model for FIFA 23 using K-Means Clustering in Python to categorize 9 million players into 6 groups; Classified players into defined groups weekly using Random Forest, enabling targeted marketing strategies for FIFA vanity releases; Utilized SQL to query 8+ KPIs measuring FIFA player engagement, monetization, and churn, delivering 6 ad hoc reports for stakeholders
Education
Hire Matthew Erxleben today
Post a job and invite Matthew, or contact them directly. No hidden commission.







