Summary
Spyridon Mouselinos is a machine-learning engineer with nine years of experience and a PhD in visual reasoning, currently building foundational ML systems at Moonvalley. He specializes in productionizing vision-language research into reliable, high-throughput platforms across media, fintech, and enterprise SaaS, having shipped multimodal Transformers and spatiotemporal GNNs that operate at tens of millions of predictions per day. His work blends rigorous research—published at ICLR, ACL and EMNLP and focused on adversarial frameworks and bias detection—with hands-on platform engineering: low-latency Kafka/Arrow feature stores, Ray/PyTorch serving, MLflow-driven CI/CD, and blue-green releases. Notably, he has architected real-time safety layers and RL-optimized trading agents with hot-swap Monte Carlo evaluation, demonstrating a rare ability to align model innovation with operational risk controls and product teams. Based in the Netherlands, he pairs deep technical breadth (graph theory, optimization, and agentic reasoning) with a practitioner’s focus on measurable business outcomes.
9 years of coding experience
7 years of employment as a software developer
Master's degree, Data Science and Machine Learning, Master's degree, Data Science and Machine Learning at Athens University of Economics and Business
GSSAS, Graph Theory and Optimization, GSSAS, Graph Theory and Optimization at Gdańsk University of Technology
Doctor of Philosophy - PhD, Visual Reasoning, Doctor of Philosophy - PhD, Visual Reasoning at University of Warsaw
Master's degree, Electrical and Computer Engineering, Master's degree, Electrical and Computer Engineering at National Technical University of Athens
English, German, Greek