Shahab Nabavi

I work in machine learning. I try not to make that my whole personality.

About

I've loved computers since I was a kid. It started with installing software, fixing small hardware problems, then learning C++ and getting into RoboCup 2D soccer simulation. At university I chose computer science because I wanted to get really good at it. The further I went, the clearer it became that becoming an expert is a lot harder than it sounds.

I did my master's at the University of Manitoba under Prof. Yang Wang. I also spent a year in a PhD program at York University before leaving to go into industry.

I started at Huawei Canada in 2020 as an R&D engineer, working on machine learning products that hadn't really existed before. Then the LLM boom hit and everything turned into a race. I didn't feel especially distinct in that landscape, so I dug in and studied it more seriously. In 2024 I joined Cerebras Systems to focus more closely on LLMs, including model bring-up, evaluation, and related work.

Around the same time I started building AI tools for internal use, especially for repetitive work people were still doing by hand. That effort went somewhere: we launched an internal product that the company still uses today.

I wasn't alone in any of this. Great colleagues have always been a big part of the work, whether through mentorship, research, code, or day-to-day collaboration.

Somewhere along the way I realized I've always had a taste for film, and a lot of dreams that might still come true someday. I've also always liked thinking philosophically. So I explored film and theatre a bit to see where I might fit. Writing turned out to be the best match. I do screenwriting part time, and I watch movies both to enjoy them and to learn how they work from another angle.

You can find me on GitHub, LinkedIn, and Google Scholar.

Recent writing

A few recent posts

Some are about AI. Some are more personal. Some are about film, theatre, science, and the future.

All posts →