Mathematics & Physics · University of Toronto · Rotman Commerce Certificate
Oakley Gompels, a Toronto-based Maths & Physics student with hands-on experience across operations, sourcing deals and AI. Seeking AI & operations roles for 2027.
I study Mathematics and Physics at the University of Toronto, alongside a Certificate in Business from Rotman Commerce. I care as much about how a decision plays out commercially as I do about the analysis behind it.
A physics degree trains one core skill: pulling signal out of noisy data and being honest about uncertainty. I've spent the last few years applying that same discipline outside the lab: building financial models at an investment-banking program, running operations under real-time pressure, and taking the lead on teams through messy, real-world problems where the answer isn't in the back of the book.
I move comfortably between the quantitative and the operational: rigorous with a model, but focused on execution and outcomes. That combination is what I want to bring to a career in AI and operations.
In summer 2026 I took on a procurement and AI development role, sourcing manufacturing for Gompels Healthcare. I built the automation that found and vetted suppliers at scale, then flew out to China to run quality control and negotiate the deal in person. The deal went from first search to production and shipment to the UK.
See the full story →Built a data-scraping pipeline to find candidate factories across China, then back-tested each one against independent sources to screen out trading and distribution companies and surface genuine manufacturers.
An AI agent opened first contact with each verified factory and logged every lead into a structured pipeline, so negotiations could start at scale rather than one email at a time.
Flew out for hands-on quality control and price negotiation to choose the final manufacturer, taking the deal from first contact through to production and shipment.
Gompels Healthcare · Melksham, UK
I support the company's move into a new warehouse, working across procurement, property and process. I built a constrained-integer racking-layout optimizer that increased usable storage by 28% while keeping forklift aisles clear of the building's structural pillars.
I also led the procurement of warehouse machinery from China. I built a data-scraping and AI outreach pipeline to find and verify manufacturers, then travelled to Shanghai to run quality inspections and negotiate prices in person. I took the order from first contact through to shipment.
On the property side, I drafted the heads of terms (letter of intent) for the new warehouse to reduce legal costs, and I backed it with market research on comparable properties to benchmark the price. Throughout the expansion, I prototyped AI-driven workflows that cut research and quality-checking time so decisions could be made faster.
EgaTree · Internship · Remote (New York)
I used AI tools to automate client performance metrics and graphs, which replaced a manual reporting process with one that can be repeated each cycle.
I researched the foundations of Bitcoin to inform the design of the company's proprietary market indicators. I then modelled financial time-series data with statistical and physics-inspired forecasting methods to test and refine those indicators.
Only Four Ltd
I processed and dispatched more than 200 orders a day at 99% accuracy during periods of high demand.
I identified the bottlenecks in the pick-and-pack process and prototyped improvements to them. Those changes were projected to save about 0.5 seconds per item, which adds up to roughly 500 hours a year.
University of Toronto Ski & Snowboard Club
I direct day-to-day operations for a student club of more than 400 members. I recruited and manage a team of more than 15 executives who deliver the club's trips and events.
I introduced inclusivity policies that lifted member engagement by 20% and reduced inefficiencies in how we plan.
Phi Gamma Delta
The chapter lives in a century-old charitable property, and relations with its owners had been strained for two decades. I rebuilt that trust and restored a working relationship between the chapter and the owners.
By planning around the owners' priorities, I unlocked a $70K renovation fund. I also designed and ran the operational workflows for a 45-member team, which improved completion rates, accountability and collaboration.
INSEAD
I presented case studies and ran startup simulations in cross-cultural teams. The program built my skills in innovation strategy and pitching.
InvestIN
I built financial models, ran equity valuations and presented M&A pitch decks in investment-banking simulations. The internship gave me exposure to capital markets and client advisory.
Humanity First, Toronto · Help 4 Homeless, Bath, UK
I have prepared and delivered 250 food and toiletry packages across Toronto with Humanity First. In Bath, I support Help 4 Homeless with ongoing meal and clothing outreach.
Lifeguard · Smallpeice Engineering Trust
Crisis management and decision-making under pressure as a qualified lifeguard; hands-on prototyping, budgeting and data analysis on engineering programs alongside Bloodhound, Rolls-Royce and JLR mentors.
AI and operations run on the same core skill a physics degree trains: extracting signal from noisy data, being honest about what you don't know, and turning it into a decision. Here's that discipline applied to a real business problem, plus a few projects from the lab and how this site was built.
For my family's business, a constrained-integer optimizer that packs the most storage racking into a 132 m × 108 m warehouse while keeping every steel pillar out of the forklift aisles. A dynamic program sweeps the floor on a 5 cm grid, trading racking gained against steel removed, reaching 29 rows at 47.5% floor-fill, within 1% of the theoretical ceiling, and reducing a messy constraint problem to one clear decision for management.
Open the interactive tool →A collection of my code, including AI automation tools built for real business problems, projects from my Mathematics and Physics coursework, and independent builds exploring areas that interest me.
View on GitHub →Modelled the stochastic motion of microspheres from video microscopy using mean-squared displacement, weighted linear fits and the Stokes–Einstein relation. A study in extracting a clean signal, and a rigorous estimate of its uncertainty, from noisy real-world data.
Read Full Report →Used the Debye–Sears effect to measure the speed of sound in water and derive its bulk modulus via optical diffraction from an acoustically driven standing wave, calibrated against a sodium lamp with careful propagation of measurement error.
Read Full Report →Investigated the departure of a real pendulum from simple-harmonic idealization, testing where a clean theoretical model breaks down against experimental data, and quantifying the systematic error in the small-angle assumption.
Read Full Report →A visual history of this website, with every design iteration preserved as a named Git branch. Explore the commits, decisions and what changed between versions.
Explore History →University of Toronto
Relevant coursework: Ordinary & Partial Differential Equations, Linear Algebra, Multivariable Calculus, Complex Variables, Combinatorics, Computational Physics, Advanced Classical Mechanics.
Supplementary: Economics, Business Ethics, Environmental Sustainability and Psychology.
Certificate in Business, Rotman Commerce: Finance, Accounting, Marketing and Business.
Dauntsey's School
A Levels: Mathematics (A), Physics (B), Economics (B).
Captain of 2nd XI Hockey; Devizes Club 1st Team Hockey; 1st Team Rugby & Cricket.
The same attention to detail I bring to my work, I bring to a photograph. What began as a hobby has become a discipline in patience, learning and iteration: noticing the moment before it happens, then refining it in post. It balances the analytical work, and I think both are better for it.
