Skills
Focused on AI — applications, safety and interpretability — built on a deep foundation of optimisation, applied mathematics and quantitative modelling.
AI & machine learning
- AI safety & scalable oversight: recursive debate protocols (honest vs. dishonest debaters), construction and analysis of obfuscated arguments, detection of subtly false sub-claims, adversarial / red-team evaluation.
- LLMs & applied AI: LLM-API pipelines; programmatic generation and analysis of structured artefacts; prompt design; automated LaTeX / PDF generation.
- Foundations underpinning ML: convex & non-convex optimisation, semidefinite programming, sum-of-squares / convex relaxations, advanced linear algebra — the mathematics behind much of modern machine learning.
- Active interests: mechanistic interpretability; trustworthy and verifiable AI.
Optimisation & applied mathematics
- Convex optimisation; relaxation of non-convex / NP-hard problems; semidefinite and second-order cone programming; sum-of-squares decompositions; interior-point methods.
- Numerical methods: Monte Carlo, finite-difference methods; stochastic differential equations.
- Probability, statistics and stochastic calculus.
Programming & software
- Languages: Python, C++, C#, SQL, VBA.
- Environments & tooling: Visual Studio, VS Code; Git, GitHub, Bitbucket, JIRA; LaTeX; Linux, macOS, Windows.
Quantitative finance
- Model validation: first-time validations, re-validations and model-change workflows; calibration, benchmarking, PAA, stress testing, risks-not-in-model.
- Pricing & models: stochastic & local volatility (Gyöngy), Linear/Quadratic Gaussian Markov rate models, Monte Carlo & PDE methods; equity & equity-rates hybrids; exotics.
- Counterparty credit risk: SA-CCR, IMM, PFE simulation, CVA; CCAR stress testing; regulatory frameworks (CRR, Basel/IOSCO).
Scientific & computational modelling
- Molecular dynamics; dissipative particle dynamics; coarse-grained simulation.
- Finite element analysis (FEA) and computational fluid dynamics (CFD).
Earlier domain & leadership
- Medical-device R&D and regulatory pathways (FDA IDE/PMA, CE Mark); IP / freedom-to-operate strategy.
- Building and leading R&D teams from the ground up (start-up first employee); project management of complex technical programmes; strong written and oral communication.