About me

Hi 👋 I’m Emmett. I like to get my hands dirty with machine learning pipelines.

Previously, I worked at Trellis AI where I deployed pipelines to extract and validate data from complex documents using large language models. I also worked as an algorithm developer at FyeLabs, where I trained and shipped custom machine learning models for industry clients. Before that, I worked as an associate at Fundata Canada, integrating computer vision into their investment fund data pipeline.

I did my M.Sc. with James Wadsley’s group at McMaster University, where I studied the capabilities of deep learning models such as CNNs, GCNNs, and Transformers for simulating galaxy-scale plasma physics.

I also love sharing my code with others. Check out thepi.pe, an open source data extraction service built with vision-language models, and engshell, an LLM-powered shell for any operating system (it’s under 100 lines of code!).

Need help? Feel free to get in touch.