Michael Mesarcik

Machine Learning Researcher & Engineer

Amsterdam · GitHub · LinkedIn · Scholar

About

ML researcher with a PhD and experience building production-grade time-series foundation models for forecasting and anomaly detection at scale. I design and deploy real-time ML systems on high-throughput data — from prototyping to production — across energy and radio astronomy.

Experience

Machine Learning Researcher — Alliander
Time-series foundation models for electric grids. 10% improvement in short-term load forecasting. Secured €420K research grant + €100K EuroHPC compute grant.
HPC Teaching Assistant — University of Amsterdam
GPU computing (CUDA, OpenMP), profiling and scaling on HPC clusters.
Data Science Intern — DataProphet
Clustering and monitoring pipelines for industrial sensor data; containerised AWS workflows.
Radar Signal Processing Intern — University of Pardubice
Drone detection algorithms for pulse-Doppler surveillance radar.

Education

PhD, Informatics — University of Amsterdam
Anomaly detection for radio telescopes at scale using machine learning. ↗ thesis
MSc, Electrical Engineering — University of Cape Town
Signal processing & remote sensing
BSc, Computer Engineering — University of Cape Town
Embedded systems & digital signal processing

Publications

ROAD: The Radio Observatory Anomaly Detector
Astronomy & Astrophysics, 2023

Projects

Skills

PyTorchTensorFlowPythonCUDAHPC / SLURMAWSGCPDockerMLOpsTime seriesAnomaly detectionGenerative models