About
I am an Applied Scientist at Amazon in Seattle, USA. Previously, I was a PhD student at McGill University, conducting my research at MILA and ILLS with Prof. Mark Coates. My work spans Bayesian and online learning, probabilistic reasoning with large language models, and graph-based representation learning, with a particular focus on controllable test-time compute and reliability.
Industry
In 2024 and 2025 I completed two Applied Scientist internships at Amazon (Seattle), where I built multi-modal pricing models for 100M+ items, introduced epistemic uncertainty quantification techniques, and delivered systems that were deployed into production. As of April 2026 I am now back at Amazon full time.
I have also collaborated with Unity Labs on motion animation in the Unity game engine, and with Quantum Technologies on recommender systems and graph learning research.
Entrepreneurship
I am Co-Founder and CTO of Isidor, where we develop vertically integrated pipelines for post training foundational models in challenging business tasks.
I am an angel seed investor in the YC startup Sourcebot.
Teaching
Teaching and mentorship have been a major part of my work at McGill. In Fall 2025 I was promoted to course lecturer for ECSE 316 (Signals and Networks). I have served as a TA and lab instructor across multiple semesters for ECSE 316, ECSE 416 (Intro to Telecommunications Systems), and ECSE 211 (Design Principles & Methods), winning the best teaching award in the Faculty of Engineering among over 200 courses offered by the Faculty.
I also designed and taught the student-led MAIS 202 — Intro to Machine Learning seminar with the McGill AI Society during my undergrad.
Before All That
I grew up in Athens, Greece, where I won a Silver Medal in the Greek National Youth Math Olympiad at the age of 15, awarded by the Hellenic Mathematical Society in 2012. Check out my 2011-2012 exam. That early love for problem-solving eventually brought me to the research I do today.