Leading a team of Data Scientists in developing ML/AI solutions for various lines of business such as Credit Risk, Fraud, Capital Markets, and T&O.

Developed machine learning based capabilities within the Enterprise Data Science Solutions Team for use cases such as insider threat detection, account takeover detection, credit resiliency modelling, pricing elasticity estimation, and credit reporting anomaly detection.

Performed 6-month rotations in the following teams:• Machine Learning Model Validation: Performed independent validations andbenchmarking of ML/AI models used within various lines of business.• Global Strategy: Developed quantitative analytics tools in Python with an Excel interface to support a team of strategists on the trading floor, e.g., a backtesting system for systematic trading, a multi-factor scorecard portfolio builder, and FX fair valuation econometric models.• Financial Engineering Model Validation: Developed C++ benchmarks to validate models such as redeemable term deposits, mortgage commitments, commodity swaptions, volatility surface generators, mortgage-backed securities, credit default swaps, and forward rate agreements.• Model Development: Developed stress testing models for financial variables (e.g., MBS OAS) to support programs such as CCAR, DFAST and OSFI-MST. Created an R package to consolidate the development and implementation of stress testing linear regression models.

• Worked on projects that aim to develop dynamic hedging strategies for segregated funds taking basis risk into account. • Worked on projects related to catastrophe (CAT) bonds; developed statistical models for forecasting the risks of CAT bond funds; and performed empirical tests of CAT bond pricing models. • Implemented a large family of regime-switching GARCH-type models in Rcpp using C++ object-oriented programming techniques (MSGARCH R package on CRAN). • Worked on a project to study the impact of parameter and model uncertainty on the forecast of risk measures. • Derived general expressions for obtaining raw, absolute, and conditional moments for a standardized version of the class of skewed distributions proposed by Fernandez and Steel (1998).

Worked on a nuclear magnetic resonance quantum processor: • Learned parameter estimation methods to characterize dipolar coupled spin systems. • Studied the different approaches to the analysis of nuclear magnetic resonance spectra. • Learned the gradient ascent pulse engineering optimal control algorithm. • Became an experimented user of computer clusters such as SHARCNET. • Worked on protocols to characterize the noise affecting the implementation of quantum gates.

Worked on an experiment to efficiently detect quantum discord: • Performed the simulations. • Wrote the tomography and discord calculation softwares. • Participated to the experiments, and analyzed experimental data.

Worked on an optical setup in which a liquid mirror, controlled via LabVIEW, automatically minimized the optical aberrations introduced by a poor quality mirror.