Maximilian
Bershtman.

Starting with the fundamentals.

In October 2021, I began my B.Sc. in Information Systems Engineering in the Technion’s Faculty of Data and Decision Sciences. My first year built the foundations in mathematics, probability, programming, and data analysis.

From systems to machine learning.

Algorithms, databases, statistics, and my first machine learning course connected the software side of my degree with the question of how models learn from data.

Alongside my studies In March 2023, I joined a deep learning research project in the Faculty of Chemical Engineering, working with graph neural networks.

Putting ideas into research.

I expanded into probabilistic models, artificial intelligence, and distributed systems. Working with research datasets gave me a practical view of how modeling choices connect to real scientific questions.

Alongside my studies My graph neural network work continued until March 2024. In January 2024, I also began a data analysis research role in the Faculty of Data and Decision Sciences.

A broader view of intelligent systems.

Advanced machine learning, causal inference, and image processing deepened my technical interests. My final year also brought together systems design and a wider perspective on how people use and interpret data.

A milestone I completed my B.Sc. in October 2025. My data analysis research role concluded in January 2025.

A new chapter in learning.

In October 2025, I started my M.Sc. in Electrical Engineering at the Technion. My focus is machine learning, deep learning, and data science, with projects spanning pretrained models and 3D computer vision.

Finding new uses for pretrained models.

My current work explores the hidden capabilities of prior-data fitted networks. With Niv Cohen, I developed creative methods for tabular anomaly detection that achieve state-of-the-art performance in our benchmark evaluation. My first paper is under review at ICLR 2027.