I'm trying to optimize a notoriously complex loss function, navigating a high-dimensional, non-convex landscape—one small step at a time. Probably, so are you!
My learnings from this winding journey:
- Choose your objective with caution—you drift toward whatever you optimize.
- Curate what you train on—you learn from what you keep feeding yourself.
- Set the pace—bold enough to move, small enough not to overshoot.
- Favour signal over noise—small, informative steps beat thrashing.
- Rein in the extremes—when gradients explode, nobody benefits.
- Beware tunnel vision (overfitting)—a little regularization keeps you adaptable.
- Use momentum—keep going; let earlier progress carry you through flat stretches.
I hold an MEng in Electrical and Computer Engineering (2015, top 3% of my class) from Aristotle University of Thessaloniki, Greece, and an MSc in Artificial Intelligence (2022, magna cum laude) from KU Leuven, Belgium. I was fortunate to learn from outstanding professors and mentors in mathematics, physics, engineering, and computer science.
I have authored two papers:
- Antoniadis, I., Vercruyssen, V. and Davis, J. (2022). Systematic Evaluation of CASH Search Strategies for Unsupervised Anomaly Detection. Proceedings of the Fourth International Workshop on Learning with Imbalanced Domains: Theory and Applications, in Proceedings of Machine Learning Research 183:8–22.
- I. I. Antoniadis, K. C. Chatzidimitriou and A. L. Symeonidis, "Security and Privacy for Smart Meters: A Data-Driven Mapping Study," 2019 IEEE PES Innovative Smart Grid Technologies Europe (ISGT-Europe), Bucharest, Romania, 2019, pp. 1–5, doi: 10.1109/ISGTEurope.2019.8905611.
My professional career started in 2015, when I joined the Centre for Research and Technology Hellas (CERTH) as a research associate (Nov 2015–Jul 2016). There, I contributed to an EU-funded H2020 project on cloud computing—when the field was still in its early stages—and worked with a large consortium of European institutions.
I continued as a software engineer at Veltio (Dec 2016–Jul 2018), an Oracle partner offering supply-chain automation solutions. I worked on real-world, large-scale problems alongside an exceptional team and led the development of data pipelines and systems used by major international retailers, including Sainsbury's in the UK.
I then joined the Intelligent Systems and Software Engineering Lab (ISSEL) at the Department of Electrical and Computer Engineering, AUTH, as an ML research engineer (Oct 2018–Sep 2021). I was the technical lead on an EU-funded project on energy monitoring and load disaggregation: applied ML research, NLP pipelines (e.g. BERT, topic models), and a high-throughput event streaming engine for real-time smart-meter analytics.
In September 2022 I joined Expedia Group in London as a machine learning scientist. On the Content & Relevance team I work on large-scale ranking and retrieval—reviews, amenities, and property understanding—using deep learning, LLMs, and multimodal methods. My recent work has included cross-brand review ranking, semantic relevance and distillation for low-latency embeddings, LLM-as-judge labelling, internal TensorFlow ranking frameworks shared across teams, distributed evaluation tooling, and research on bias, calibration, and data pruning.
*My first job was in 2011, during my second year at AUTH, as a part-time support representative at OTE, the largest telecommunications company in Greece.
I find it exciting to push human boundaries with technology, and I believe we have a responsibility to leave the world better for future generations.
All it takes is one small step at a time!
Past side projects, coursework, and research code live in separate repositories. Most are archived on GitHub (read-only snapshots; not actively maintained).
Personal notebooks & experiments (self-directed; not part of a degree curriculum)
- bias-variance-decomposition — bias–variance decomposition
- fair-binary-classification — fairness on Adult (AIF360)
- gaussian-bandits — multi-armed bandits
- example-level-gradient-analysis — per-example gradients
- kepler-exoplanet-prediction — Kepler / KOI classification notebook
- sorting — Python sorting algorithms, pytest, benchmarks
KU Leuven — Master of Artificial Intelligence
Coursework
- cart — CART / decision trees
- grid-world-mdp — grid-world MDP, policy iteration
- taxi-rides-mapreduce — Hadoop MapReduce & Spark taxi analytics
- locality-sensitive-hashing — LSH on Stack Overflow posts
Thesis & published research
- cash-for-unsupervised-ad — Master’s thesis code extended to the LIDTA 2022 (ECML/PKDD) paper: CASH / AutoML for unsupervised anomaly detection
Aristotle University
Diploma thesis
- insight-qa — Semantic question answering (Java, Elasticsearch, LDA)
Coursework
- pagerank, octree-division — Parallel C (PageRank; octree spatial division)
You can find my full CV here.