I am a data scientist and applied mathematician with a Ph.D. in mathematics, working at the intersection of machine learning, scientific AI, network science, and data-driven decision support.

My current work focuses on building practical machine-learning systems for complex scientific and industrial data. I am especially interested in roles where statistical thinking, model evaluation, optimization, and clear communication are needed to turn limited or noisy data into useful decisions.

Contact: konstantinos_pantazis@hotmail.com	+30 695 5532 846	Athens, Greece
Profiles: LinkedIn	Google Scholar	ResearchGate

CVs: Industry CV

Academic CV

Sections: Industry

Academia

Personal

Industry

I am currently a Data Scientist at Deus Ex Machina, a subsidiary of PIPA AI, where I work on machine-learning and data-science systems for scientific and industrial applications.

My industry profile is centered on:

Machine-learning pipelines for heterogeneous scientific and operational datasets.
Predictive modeling, model evaluation, feature engineering, and optimization.
Scientific AI and decision-support workflows for complex real-world systems.
Data integration from APIs, databases, public sources, and structured scientific records.
LLM, NLP, and retrieval workflows for scientific and biomedical information extraction.
Translating technical results into clear reports and stakeholder-facing recommendations.

Selected Industry Experience

Data Scientist, Deus Ex Machina / PIPA AI
Athens, Greece / Remote, September 2024 - Present

Build machine-learning workflows for scientific and industrial datasets, including predictive modeling, feature selection, automated retraining, and model evaluation.
Develop decision-support workflows that combine data processing, statistical modeling, and optimization.
Work with heterogeneous data sources, including APIs, databases, public datasets, and structured scientific information.
Evaluate LLM, NLP, and retrieval-based systems for information extraction and scientific knowledge workflows.

Joint E+D & MSR Research Intern, Microsoft
Seattle, USA, May 2022 - August 2022

Modeled security incidents as graphs and applied graph-similarity learning methods for relational incident analysis.
Processed network-streaming security data and built interactive Power BI dashboards for exploratory analysis and risk quantification.

Machine Learning Intern, DEVCOM Army Research Laboratory
Washington, DC, USA, June 2021 - August 2021

Surveyed recommender-system methods and proposed collaborative filtering approaches for prioritizing information objects in dynamic, high-volume settings.

Technical Areas

Programming and data systems: Python, SQL, R, MATLAB, Power BI, REST APIs, scikit-learn, PyTorch, TensorFlow, Polars, NetworkX.
Machine learning and analytics: regression, classification, clustering, feature engineering, feature selection, model evaluation, graph learning, recommender systems.
Scientific AI and optimization: surrogate modeling, model calibration, active learning, uncertainty-aware evaluation, statistical inference, process optimization.
LLMs and NLP: HuggingFace, BioBERT, LangChain, RAG, RAGAS, TruLens, Pydantic, LLM APIs, graph-based retrieval.

For a complete industry-focused summary, please see my Industry CV.

Academia

My academic background is in applied mathematics, statistical network analysis, graph matching, multiplex networks, and machine learning for heterogeneous relational data. My research develops methods for inference across multiple graphs, entity alignment, and joint network embeddings, with applications including neuroscience, social platforms, security incident networks, and structured scientific datasets.

Education

Ph.D. in Mathematics, University of Maryland, College Park, 2019 - 2022
Dissertation: Statistical Inference across Multiple Graphs: Advancements in Multiplex Graph Matching and Joint Spectral Graph Embeddings
Advisor: Professor Vince Lyzinski

Ph.D. Studies in Mathematics, University of Massachusetts Amherst, 2017 - 2019
Transferred to the University of Maryland following advisor relocation.

B.Sc. in Applied Mathematics, National and Kapodistrian University of Athens, 2011 - 2016

Academic Experience

Postdoctoral Fellow, Applied Mathematics & Statistics
Johns Hopkins University, June 2022 - June 2023

Developed statistical methods for network-valued and time-series-derived data, with applications to neuroscience, representation learning, dependence modeling, and downstream inference.
Studied relationships between network embeddings and classical statistical techniques such as principal component analysis.
Faculty sponsor: Professor Carey Priebe.

Research Assistant
University of Massachusetts Amherst, September 2018 - August 2019

Worked on graph matching problems under the supervision of Professor Vince Lyzinski.

Publications and Preprints

Hybrid multi-physics and machine-learning framework for prediction and optimization in twin-screw food extrusion.
Eleftherios Tsivolas, Konstantinos Pantazis, Dimitrios Sfondylis, Marina Kontou, and Ilias Tagkopoulos. Journal of Food Engineering, under review, 2026.
Optimizing the Induced Correlation in Omnibus Joint Graph Embeddings.
Konstantinos Pantazis, Michael Trosset, William N. Frost, Carey E. Priebe, and Vince Lyzinski. Journal of Computational and Graphical Statistics, accepted for publication, 2026. arXiv
Clustered Graph Matching for Label Recovery and Graph Classification.
Zhirui Li, Jesus Arroyo, Konstantinos Pantazis, and Vince Lyzinski. IEEE Transactions on Network Science and Engineering, 10(6):3384-3395, 2023. Journal
The Importance of Being Correlated: Implications of Dependence in Joint Spectral Inference across Multiple Networks.
Konstantinos Pantazis, Avanti Athreya, Jesus Arroyo, William N. Frost, Evan S. Hill, and Vince Lyzinski. Journal of Machine Learning Research, 23(141):1-77, 2022. Journal
Multiplex Graph Matching Matched Filters.
Konstantinos Pantazis, Daniel L. Sussman, Youngser Park, Zhirui Li, Carey E. Priebe, and Vince Lyzinski. Applied Network Science, 7, Article 29, 2022. Journal

Selected Talks and Presentations

CMStatistics 2022, 15th International Conference of the ERCIM WG on Computational and Methodological Statistics, December 17-19, 2022. Slides
NeurIPS 2022, Thirty-sixth Conference on Neural Information Processing Systems, November 28 - December 1, 2022. Poster
Invited talk at the National and Kapodistrian University of Athens, December 10, 2021. Announcement
2021 Joint Statistical Meetings, August 2021. Video presentation
DEVCOM ARL Summer Student Symposium, August 12, 2021. Poster
NETWORKS 2021, A Joint Sunbelt and NetSci Conference, July 2021.

Teaching

Course Instructor, Johns Hopkins University

Introduction to Optimization (EN.553.361), Spring 2023.
Probability and Statistics for the Biological Sciences and Engineering (EN.553.311), Fall 2022.

Teaching Assistant, University of Maryland, College Park

Introduction to Linear Algebra (MATH240), Spring 2022.
Applied Probability and Statistics (STAT400), Spring 2020 and Spring 2021.
Linear Algebra for Scientists and Engineers (MATH461), Fall 2019.

Teaching Assistant, University of Massachusetts Amherst

Ordinary Differential Equations (MATH331), Fall 2017.

Reviewing and Service

Reviewer for the Journal of Machine Learning Research.
Reviewer for Knowledge-Based Systems.

For the full academic version, please see my Academic CV.

Personal

Outside work, I value community service, health, and time outdoors.

Blood and platelet donor in Greece and the United States.
Former probationary / associate member of Hyattsville Volunteer Fire Department, Inc..
Interested in hiking, running, cycling, climbing, and outdoor exploration. You can also find me on Strava.