At Engel Systems, I'm developing a predictive maintenance system for a moped fleet using Python and 10 Hz telemetry data. I engineered ride-level features from more than 37,000 sensor rows, including battery voltage sag and vibration levels.
I built a staged anomaly detection pipeline with Isolation Forest to flag abnormal rides, and designed a behavior-event taxonomy for mechanical stress indicators. I also developed fleet-health dashboards to track battery and vehicle-health KPIs.
At We Developers, I wrote SQL queries to extract and validate application data, and analyzed survey data with Python and Pandas. I also developed a desktop application to support data-driven decision-making.
Earlier, at Capgemini, I analyzed production data and worked with DB2 databases to validate and troubleshoot enterprise systems. I led a system optimization project that increased throughput by 150%.

