Hi, I'm Michal!

I make AI drug design happen at Grove Biopharma (Bionic Biologics™), where I am the whole engineering team: from racking GPUs and managing AWS to building data systems and shipping the apps our scientists use. I joined Grove when it acquired the ML team from Fable Therapeutics, an AI protein design startup where I led MLOps. Before that I was the first employee at Semantic Health, where I built and led the machine learning function all the way to acquisition. More detail is on my bragsheet.

Core Interests

Education - Through the incredibly hard work of my parents I was awarded the best possible education. I want everyone else to have the same opportunities I did. Whenever I have the bandwidth I share my knowledge by teaching Machine Learning / Statistics at my alma mater, the University of Toronto.
Healthcare - My interest in applying Machine Learning to healthcare was sparked early on in university days, but it really grew thanks to Prof. Marzyeh Ghassemi, who taught the ML4Health course at UofT and who made it possible for us to collaborate with the team at St. Michael's Hospital in Toronto working on documents related to Multiple Sclerosis.
BioTech / Proteins - Sparked by my time at Fable and now my day job at Grove: protein structure prediction, molecular dynamics, and making the GPUs behind them go fast.
Food - I love to cook, photograph, share, and review food across various cuisines and explore restaurants wherever I go. I run a not-so-small (27 people) "food group" in Toronto.
Agriculture - Everyone who has worked with me / is friends with me has seen my obsession with agriculture (permaculture, AI in ag, regenerative farming) grow over the past couple of years. It's just another natural extension of my passion for food.
Photography - The most recent one of my non-work hobbies, just getting started but this has been a natural progression from taking food photos.

Code

Most of my work is in Python, but I'm familiar with JavaScript / TypeScript and the "obvious" (SQL, R).

Tools

AWS, Slurm, Docker, Kubernetes, GitHub Actions, Prometheus / Grafana, PostgreSQL, uv, MLflow, W&B, and a growing set of AI coding agents.

Libraries

PyTorch, CUDA, Triton, cuEquivariance, FlashAttention, DeepSpeed, Hugging Face Transformers, Lightning, and many more.

Michal Malyska AI x bio

Hi, I'm Michal!

Core Interests

Code

Tools

Libraries