Matias Barandiaran

Matias Barandiaran

Quantitative Researcher


I do quantitative research at a proprietary trading firm. In light of recent events, I don't name the firm publicly to protect the privacy of the people I work with.

Currently living in Leipzig, Germany.

I have a background in computer science and am interested in machine learning, financial markets and speedy software. Lately, most of my research is on prediction markets.

I also enjoy playing badminton and fishwatching.

Timeline

2026 -
I am a quant researcher at a proprietary trading firm.
2024 - 2025
I was a research assistant at the Technical University of Munich, where I focused on optimizing deep learning methods for rare variant association testing. I worked with Eva Holtkamp at the Gagneur Lab and I also had the pleasure of receiving mentorship from Julien Gagneur and Oliver Stegle.
2023
I was a research intern at Charles University, in the heart of Prague's Old Town. I worked with Martin Blicha on improving the reliability of Horn solver outputs by mapping internal reasoning into verifiable proofs.
2022 - 2025
BSc at Lancaster University with a major in computer science and a minor in quantitative methods. There I became interested in developmental processes and published my dissertation on morphogenetic reservoirs with James Stovold.

Writing

I write on Bear blog. Some of my favourite posts:

All posts

Pet projects

Things I've built on the side. More on my GitHub.

graph-growing-nca is a graph-based generalization of neural cellular automata for single-cell morphogenesis.
pytorch-tutorial is an end-to-end PyTorch tutorial on supervised image classification with CIFAR-10.
mind-weaver is a 2D puzzle platformer built around line-of-sight control.
dnd-rulebook is a document extraction pipeline that parses a D&D rulebook into structured data.

Publications

Also on Google Scholar.