Research Experience
Climate-Informed Mortality Modelling for Insurance and Financial Risk
May 2025 - Aug. 2025
University of Waterloo (Waterloo, ON)
Mentor: Rhoda Dadzie-Dennis
I studied whether climate signals can improve long-horizon mortality forecasting and how this connects to insurance and financial risk. I implemented and extended the Lee–Carter mortality framework using Ontario mortality data (1991–2023), and tested a climate-augmented version by introducing mean annual temperature as a covariate. Results showed that adding temperature improved model fit and provided more decision-relevant long-term risk signals compared with a baseline Lee–Carter model, especially when forecasting under warming trends. I summarized the motivation, methodology, and implications in an editor-reviewed publication: “The Financial System in a Warming World,” Notes from the Margin (Canadian Mathematical Society), Vol. XIX (2025).
Skills/Tools: R, Python · time-series forecasting · mortality modeling · model evaluation · climate risk
Formal Theorem Proving in Applied Mathematics Using Lean 4
Sept. 2025 – Dec 2025
University of Waterloo (Waterloo, ON)
Mentor: Sita Gakkhar
I explored how formal verification can strengthen mathematical reliability by expressing and proving results in Lean 4. My work focused on optimization foundations and mathematical structures needed for linear programming. I practiced proof construction in dependent type theory and tactic-based workflows, then built formal objects (vectors, dot products, feasibility conditions) to support a machine-checked proof. As a capstone, I completed a verified proof of the Weak Duality Theorem for linear programming in Lean 4.
Skills/Tools: Lean 4 · formal verification · optimization theory · proof engineering
Mathematical Modelling of Cell–ECM Mechanotransduction
Jan. 2026 – Apr. 2026
University of Waterloo (Waterloo, ON)
Mentor: Gordon Robert McNicol
I studied how cells sense and respond to the mechanical properties of their extracellular matrix (ECM) through a process called mechanotransduction. I implemented and analysed a system of 18 coupled ODEs capturing integrin adhesion dynamics, Rho/ROCK signalling, actin polymerisation, and myosin activation. I performed parameter sensitivity sweeps on three key quantities — internal feedback strength (δ), ECM ligand density (n_s0), and ROCK inhibitor binding rate (k_IR⁺) — revealing bistable switch-like behaviour in cell adhesion commitment and identifying a critical substrate site threshold (~200 µm⁻²) below which mechanotransduction cannot initiate. I also modelled ROCK inhibitor dose-response dynamics, generating a testable prediction that inhibited cells should exhibit increased nascent adhesions but no focal adhesion maturation. Results were presented at the Directed Reading Programme symposium.
Skills/Tools: Python · MATLAB · ODE systems · parameter sensitivity analysis · mechanotransduction · mathematical biology