Resume

• Built and deployed algorithmic market-making strategies, incorporating volatility and skew estimation, order flow modelling, queue-aware execution, market impact, and reference price dynamics.
• Built a dynamic liquidity allocation model for smart order routing, reducing slippage and transaction costs by an average of 5.8% through real-time optimization over aggregated, fragmented order books.
• Developed quantitative portfolio rebalancing methods across exchanges that optimised the timing and execution of inventory transfers under latency constraints, transaction costs, and market liquidity.

• Monitored production market-making algorithms, managing real-time risk parameters and system health to minimize inventory exposure during high-volatility periods.
• Designed and deployed a live trading terminal using ccxt and native exchange APIs for real-time position and order tracking, integrating Tardis.dev for historical position reconstructions.

• Teach statistics, covering hypothesis testing, regression theory, and statistical inference.
• Supervise a group project in which students apply these methods to an economic analysis of their choice.

• Designed scalable likelihood-based estimation algorithms for functional scale models, optimising computational performance through vectorised computations and parallel processing.
• Reduced execution time of large-scale Monte Carlo simulations by over 90%.

• Implemented a statistical learning framework based on dynamic Cox proportional hazards models to estimate event probabilities and customer retention dynamics in high-dimensional data.
• Developed parameter estimation procedures and feature engineering pipelines to address right-censoring, non-stationarity, and covariate shift, improving out-of-sample prediction stability.

• Researching statistical learning theory, high-dimensional probability, and stochastic processes.
• Organized and led a weekly graduate seminar on empirical process theory and statistical learning theory, with applications to machine learning and deep learning.

The full technical writeup, including a walkthrough of the \(\gamma_2\) functional and the majorizing measure theorem, is on the Blogposts page.

• Honours Programme, GPA 8.9/10 (magna cum laude).
• Developed a novel nonparametric testing framework for multidimensional stochastic processes for a thesis project, implementing and packaging the mathematical framework into an open-source Git repository (see Projects).
• Relevant coursework: Measure Theoretic Probability, Quantitative Financial Risk Management, Stochastic Processes, Stochastic Integration.

• Applied deep learning models for binary classification of digital asset news using Keras, with text preprocessing and feature engineering.

• Extending the functional GARCH framework to a generalized autoregressive score (GAS) model to estimate and capture time-varying intraday volatility surfaces.
• Designed efficient estimation procedures using B-splines, applying Numba JIT compilation to enable scalable modelling of volatility surfaces from granular intraday return data.

• Implemented Hierarchical Risk Parity via tree clustering using scipy.cluster to stabilize high-dimensional asset allocation, bypassing classical covariance inversion to eliminate noise sensitivity.
• Outperformed the benchmark by an average of 3.9 percentage points across multiple simulated investment horizons in a look-ahead-free backtest.

• Represented the interests, opinions, and concerns of the PhD candidates.
• Contributed to improvements in budgeting, strategic planning, and day-to-day matters.

• Participated in live Q&A sessions and (virtual) educational fairs.
• Assisted prospective students by answering questions on the online chat platform.

Languages & Tools

• Python (Pandas, NumPy, Numba)
• Java
• SQL
• Git
• Bash
• Dask

Scientific Computing

• SciPy
• statsmodels
• CVXPY
• scikit-learn
• TensorFlow

Quantitative Methods

• Time series analysis
• Econometrics
• Statistical inference
• Optimisation