Data science · Software engineering · Ontario, Canada

I build clear, useful systems from complex information.

I'm a Data Science and Software Engineering student at Western University, currently working in data science at the Electrical Safety Authority.

Experience

2023 — Present
HackathonsFour events · Selected experiencesExplore

Offers

Career milestones

Projects

GitHub archive
Hack the North 2026 · Solana trackChalk ChainView repository ↗

Teacher absence runs near one in four in parts of the world, and paying teachers for days they can prove they taught is known to cut it — but the proof gets faked, because a photo carries no trustworthy timestamp. Chalk Chain turns a live Solana block into three words the teacher chalks on the board before photographing the class. A Solana program, not a server, checks the photo was sealed within 150 seconds of those words existing; surprise re-checks timed by future block hashes extend a hash chain written in chalk; a vision model reads the board, and each verified photo pays a USDC bonus that anyone can audit.

Three words drawn from a Solana block, chalked on a boardPublic proof page showing a day of verified photos and the USDC bonus paidThe teacher app on a phone, showing each photo and which checks passed9On-chain instructions175Tests, 40 end-to-end3 / 3Staged cheats rejected<$0.001To verify and pay a dayRustAnchorSolanaTypeScript@solana/kitReactPythonFastAPIVision LLM
Featured projectEV charging gap modelView repository ↗

A two-stage hurdle model that finds Canadian cities short of fast DC charging, then sizes the shortfall. A prescriptive layer turns each predicted gap into charger counts, urgency tiers and costs, and a mixed-integer solver picks the highest-impact sites that fit a fixed capital budget.

Top 20 Canadian cities ranked by predicted fast-charging deficit, led by Montréal at +165 chargersUrgency tier distribution across 326 gap cities: 45 critical, 29 high priority, 87 moderate, 165 lowSHAP summaries for the classification and regression stages, with station count the dominant feature2,982Cities scored1,032Chargers prescribed20Models per stage0.96Holdout R²RtidymodelsSHAPShinyMILP / GLPKdeck.glPython
Selected workBrowse the build archive.Data products, machine learning experiments, and software tools—all in one place.View on GitHub ↗

About

More than a résumé

My work sits where data, software, and product thinking meet. I've built data pipelines, machine learning models, dashboards, and AI tools that make information easier to understand and act on.

I'm especially interested in practical machine learning: systems that leave the notebook, reach real users, and improve how decisions get made.

Away from the screen, you'll usually find me playing badminton, lifting weights, or analyzing market trends.