Hi.

Welcome to
my Website!

I am Nikolaos :)

I am an
AI Scientist

I ♥ building models for
scientific discovery.

About Me

I'm currently an AI Scientist, building and applying machine learning methods in biology and clinical science. My path has taken me from chemical engineering at NTUA, to applied mathematics and dynamical systems at JHU, and now to industry research at Genentech—evolving from classical simulations to data-driven modeling and, most recently, agentic AI. Throughout this journey, the driving force has been a passion for leveraging computation to accelerate scientific discovery.

Roots in Greece & Cyprus

Origins

I was born in Cyprus and grew up between Greece and Cyprus—two places that shaped my sense of identity, community, and my appreciation for good food.

Academic Projects

Showing 18 projects

A Mechanistic Analysis of Transformers for Dynamical Systems

A Mechanistic Analysis of Transformers for Dynamical Systems

Transformers for Dynamical Systems Mechanistic Interpretability Transformers' Interpretability

Enabling Probabilistic Learning on Manifolds through Double Diffusion Maps

Enabling Probabilistic Learning on Manifolds through Double Diffusion Maps

Probabilistic Learning on Manifolds Generative Modeling Geometric Harmonics

Comparing analytic and data-driven approaches to parameter identifiability: A power systems case study

Comparing analytic and data-driven approaches to parameter identifiability: A power systems case study

Parameter IdentifiabilityPower SystemsManifold Boundary Approximation Method (MBAM)

Generative Learning for Slow Manifolds and Bifurcation Diagrams

Generative Learning for Slow Manifolds and Bifurcation Diagrams

Generative ModelingBifurcation Discovery

Machine learning for the identification of phase transitions in interacting agent-based systems: A Desai–Zwanzig example

Machine learning for the identification of phase transitions in interacting agent-based systems: A Desai–Zwanzig example

Phase Transition LearningCollective DynamicsRepresentation Learning

Tipping Points of Evolving Epidemiological Networks: Machine Learning-Assisted, Data-Driven Effective Modeling

Tipping Points of Evolving Epidemiological Networks: Machine Learning-Assisted, Data-Driven Effective Modeling

Network EpidemiologyPhase Transition LearningData-Driven Disease Modeling

Task-oriented machine learning surrogates for tipping points of agent-based models

Task-oriented machine learning surrogates for tipping points of agent-based models

Digital TwinsPhase Transition LearningNeural SDEs

Nonlinear dimensionality reduction then and now: AIMs for dissipative PDEs in the ML era

Nonlinear dimensionality reduction then and now: AIMs for dissipative PDEs in the ML era

Representation LearningPDE SurrogatesApproximate Inertial Manifolds

Active learning of ternary alloy structures and energies

Active learning of ternary alloy structures and energies

Active LearningBayesian OptimizationAI for Material Discovery

From Noise to Signal: Unveiling Treatment Effects from Digital Health Data through Pharmacology-Informed Neural-SDE

From Noise to Signal: Unveiling Treatment Effects from Digital Health Data through Pharmacology-Informed Neural-SDE

Neural-SDEsAI for PharmacologyDigital Health AI

Double diffusion maps and their latent harmonics for scientific computations in latent space

Double diffusion maps and their latent harmonics for scientific computations in latent space

Kernel Based Regression (Geometric Harmonics)Surrogate ModelsRepresentation Learning

Machine learning-assisted crystal engineering of a zeolite

Machine learning-assisted crystal engineering of a zeolite

Zeolite Inverse DesignAI for Material Science

On the Parameter Combinations That Matter and Those That Do Not: Data-driven studies of parameter (non)identifiability

On the Parameter Combinations That Matter and Those That Do Not: Data-driven studies of parameter (non)identifiability

Parameter IdentifiabilityDiscovery of Sensitivity LandscapesRepresentation Learning

From partial data to out-of-sample parameter and observation estimation with Diffusion Maps and Geometric Harmonics

From partial data to out-of-sample parameter and observation estimation with Diffusion Maps and Geometric Harmonics

Diffusion ManifoldsOut-of-Sample EstimationKernel-Based Learning

Learning effective SDEs from Brownian dynamic simulations of colloidal particles

Learning effective SDEs from Brownian dynamic simulations of colloidal particles

Coarse-Grained SimulationNeural SDEsData-Driven surrogates

Learning effective stochastic differential equations from microscopic simulations: combining stochastic numerics and deep learning

Learning effective stochastic differential equations from microscopic simulations: combining stochastic numerics and deep learning

Neural-SDEsDynamics IdentifiabilityGenerative Scientific AI

Initializing LSTM internal states via manifold learning

Initializing LSTM internal states via manifold learning

LSTMsTemporal Representation Learning

Coarse-grained simulations of bidisperse polymer melts

Coarse-grained simulations of bidisperse polymer melts

Coarse-grained simulationsBrownian dynamics/kinetic Monte Carlo

Personal Experimentations

Building an agentic wrapper for scigma

Building an agentic wrapper for scigma

Agentic AILangchainLarge Language Model

Test Time Compute Experiments

Test Time Compute Experiments

Inference Time ComputeLarge Language ModelsBag of NReflection

Testing TabPFN on Zeolitic Data

Testing TabPFN on Zeolitic Data

Foundation ModelsTabular data

Assessing Kosmos on Zeolites

Assessing Kosmos on Zeolites

Agentic AIScientific Discovery

Assessing Kosmos on Coarse Bifurcation Diagram Creation

Assessing Kosmos on Coarse Bifurcation Diagram Creation

Agentic AIPhase Transition LearningCoarse Bifurcations

LoRA Experimentations

LoRA Experimentations

Low Rank ApproximationLarge Language ModelsFine-tuning

Ask Nikolas RAG

Ask Nikolas RAG

Large Language ModelsRetrieval Augmented Generation

Reinforcement Learning for Toy Systems

Reinforcement Learning for Toy Systems

Reinforcement LearningReinforceProximal Policy Optimization

Contact Me

Contact Visual