I'm a Data Engineer at Commsquare, building real-time Spark/Kafka pipelines that process 100M+ events/day for a top EU telecom provider. Before that, I was a Research Assistant at Athena Research Center, where RL-based demand forecasting models I built achieved a 33% efficiency improvement on cloud infrastructure — results published in peer-reviewed journals. And before that, a Data Scientist & ML Engineer at Meteorologica, where deep learning models I designed improved wind energy forecast accuracy by 23%.
I hold a 5-year MEng in Electrical & Computer Engineering from NTUA Athens (CS major, thesis: Emotion Driven Speaker Verification). I'm a Google Summer of Code alumnus — my document anonymization system was presented at the Hellenic Council of State. I'm a published researcher, and I tutored Computer Vision at the M2L Summer School 2023.
I build things that process data at scale — and lately, AI systems that think for themselves.
Designed and implemented a configurable KPI engine used in production — enables on-demand analytics pipelines with reduced feature deployment time. Optimized and maintained real-time data pipelines with Spark Structured Streaming + Kafka, processing hundreds of millions of events/day for a top EU telecom provider.
Demand forecasting pipeline combining RL + statistical models for data center energy optimization — 33% efficiency improvement (tested on cloud infrastructure). Co-authored peer-reviewed publications in international journals and conferences. Supervised undergraduate and graduate students. Collaborated under the ARCADIA project with Intracom.
Deep learning architectures for time-series forecasting — 23% forecast accuracy improvement for energy production. End-to-end design and maintenance of production pipeline on Microsoft Azure. Daily energy forecasts for wind farms; clients: Terna and Inaccess.
Feature extraction methods for audio tasks. Developed an open-source PyTorch library for training, testing, and fine-tuning audio models.
Built a complete system for anonymizing sensitive documents — web UI + LibreOffice extension. Presented at the Hellenic Council of State (Supreme Administrative Court of Greece).
5-year joint degree, 300 ECTS, major in Computer Science. Thesis: Emotion Driven Speaker Verification. Coursework: Machine Learning, Pattern Recognition, NLP, Algorithms & Complexity, Software Engineering, Computer Networks, Operating Systems.
Tutor at the Mediterranean Machine Learning Summer School. Organized the Computer Vision segment with daily tutorial sessions.
Top 10 score in Greece. Built a routing algorithm for self-driving vehicles.
Co-authored peer-reviewed publications at Athena Research Center on RL-based demand forecasting for data centers.
Greek (native), English (Michigan Proficiency), German (B2, Goethe-Institut).