I'm currently a Senior Data Scientist at BioSymetrics.
Tech moves fast, even in the drug development space.
I introduced modern tooling, including Snakemake, uv, and Pixi, to BioSymetrics, allowing us to move quickly on code changes and dependency updates.
I also introduced AI tooling, including OpenAI Codex, to BioSymetrics, substantially reducing the time needed for data science explorations and feature work.
We encountered a dataset of 550 million compounds.
I trained a chemical foundation transformer model, which remains a core asset under active development at BioSymetrics.
Researchers from the Parent Project Muscular Dystrophy consortium came to us with a clinical and proteomic dataset.
I analyzed this dataset using machine learning, identifying biomarkers of disease trajectory for muscular dystrophy.
These efforts are published in Nature Communications.
Our parent company underwent a rebrand.
With some tasteful AI help, I generated new visual assets and collaborated with an external marketing and design agency on our parent company's website.
I also established and expanded our parent company's social media presence to LinkedIn, X, Bluesky, Threads, and Instagram.
Once upon a time, I did my Ph.D. in Immunology at the University of Toronto and SickKids, and then stayed on as a Bioinformatician.
Childhood arthritis had a suboptimal classification system, and we had multi-omic and clinical data that could help us classify better.
I used unsupervised machine learning to identify five data-driven subtypes, published in Arthritis Rheumatology.
I then used a different, novel application of unsupervised machine learning to identify seven data-driven patterns of joint involvement, published in PLOS Medicine.
These findings formed a foundation for the Understanding Childhood Arthritis Network (UCAN), an active consortium of Canadian and Dutch sites.
I did computational work in a traditionally non-computational field.
I refined the art of communicating concepts to non-expert audiences across various disciplines through poster and podium presentations.
I mentored medical school students and residents on using machine learning in their areas of expertise, with two publications in ACR Open Rheumatology and another in Arthritis Rheumatology.
My transition to BioSymetrics would have left substantial personnel gaps in my group at SickKids.
I moved to an advisory role, helping to hire several candidates to fulfill my previous full-time responsibilities.
I still advise for my previous group at SickKids, providing feedback into their analyses and sharing the latest developments in AI and machine learning.
But wait, there's more!
The Immunology Graduate Students Association (IGSA) at the University of Toronto was bound to a constitution with ≥11-year old revisions that no longer reflected how the association operates. This would've left us in violation with the Graduate Student Union guidelines.
As IGSA's co-president, I spearheaded the overhaul of its constitution, leading a process of public consultation. The new constitution was unanimously ratified.
I started as a soundboard operator at Grace Toronto Church, but noticed we had a shortage of bass guitarists.
I learned the bass guitar, leaning on my past experiences in other musical endeavours and elective studies in music theory during my undergraduate studies. And while I now only infrequently step behind the soundboard, my experiences on the bass guitar and worship team gives me a unique perspective both behind the soundboard and when advising other soundboard operators.