ABOUT

I'm a strategic risk & data leader in global energy markets with a decade of experience turning complex commodity data into decisive trading and investment action.

WHAT I DO

→
Risk Modelling: VaR & liquidity models, Monte-Carlo simulations, scenario analysis, initial-margin calculations
→
Quantitative Analysis: PPA pricing models, gas storage optimisation, hedging strategies, capital allocation
→
Data Engineering: Full-stack data platforms, real-time streaming (Kafka), cloud infrastructure (AWS/Azure)
→
Machine Learning: Demand forecasting using ANNs, time-series analysis (SARIMAX), TensorFlow/PyTorch implementations
→
Team Leadership: Leading cross-functional teams, Python upskilling programmes, stakeholder management

CURRENT FOCUS

→
Building AI agents for commodity market analysis and trading strategy generation
→
Optimising LLM workflows for financial data extraction from market reports and forecasts
→
Creating production-ready systems that combine traditional quant models with modern AI capabilities
→
Exploring token optimisation techniques for processing high-frequency commodity data
→
Testing AI claims through rigorous experimentation on real energy market data

WHY I WRITE

I believe in testing claims, not just accepting them. When I see a new AI tool or technique promising dramatic improvements, I build tests, run experiments, and share what I learn. My goal is to help others make informed engineering decisions based on real data, not marketing hype.

EXPERIENCE

A decade in global energy markets spanning risk management, quantitative analysis, data engineering, and algorithmic trading. I've built production risk systems, led technical teams, and traded commodities across Europe and the Middle East.

FEATURED PROJECT

TOON vs CSV vs JSON Analysis

Comprehensive testing of token optimisation claims on real financial data. Includes test methodology, code, and analysis.

VIEW ON GITHUB →

GET IN TOUCH

I'm always interested in connecting with fellow developers, researchers, and engineers working on similar problems.

EMAILLINKEDINX/TWITTERGITHUB