Senior AI/ML scientist and engineer building reliable production systems, with a quantitative-markets research background.
I work across applied research, modelling, deployment, evaluation and production AI infrastructure. My focus is the boundary between model capability and operational reliability: systems that are measurable, auditable and safe to run at scale.
| Project | What it demonstrates |
|---|---|
| CatalystLens | An evidence-checked financial catalyst research application with FastAPI, optional local LoRA inference, citation validation, abstention, observability, CI and a live demo. |
| Quant Regime Engine | A deterministic, independently tested factor and regime engine designed to remain auditable when called by an AI agent. |
| Cross-Asset Shock Diffusion | Leakage-aware quantitative research with chronological validation, next-open execution, transaction costs, falsification tests and reproducible reports. Paper and DOI. |
These repositories use public or synthetic data and state their assumptions, evaluation boundaries and limitations. They demonstrate engineering and research methods rather than investment performance.
- Production AI: LLM and agentic applications, retrieval, evaluation, inference optimisation, observability and lifecycle controls
- Applied ML: forecasting, ranking, recommendation, experimentation, calibration and feature engineering
- ML platforms: scalable inference and data pipelines, deployment, monitoring and reliability
- Quantitative research: temporal validation, market regimes, transaction costs, risk and reproducibility
I publish evidence-led work on AI, machine learning, macro and quantitative markets.
Open to senior AI/ML engineering, applied science and quantitative-technology opportunities.