I'm a quantitative analyst working at the intersection of credit risk modelling, data analytics and technology.
My professional background is in banking and consulting, with a focus on IRB and IFRS 9 model development across PD, LGD and EAD for Retail, Retail SME and Wholesale portfolios. Outside of my core credit risk work, I build full-stack and serverless applications using Python, Vite, React, TypeScript and AWS.
I specialise in credit risk modelling across secured and unsecured lending products, with experience in:
- IRB and IFRS 9 model development
- PD, LGD and EAD modelling
- Model monitoring and validation support
- Regulatory remediation and inspection support
- Stakeholder engagement and technical delivery
I'm particularly interested in practical tools that combine quantitative modelling, clean data workflows, cloud architecture and accessible user interfaces.
Interactive tool for building credit risk scorecards from raw data - covers the full development pipeline from factor screening and WoE/IV analysis through logistic regression, PDO scaling, stability assessment and Excel report export.
Reference implementation and interactive demo of the Monotone Adjacent Pooling Algorithm for score-to-PD calibration, with side-by-side implementations in Python, C++, R, MATLAB and SAS.
Six-step wizard that walks through an LGD model-development pipeline - upload a raw monthly loan-panel, construct default episodes and recovery cash flows, calculate Loss Given Default under workout, market and implied-market methodologies, review vintage and stability diagnostics, calibrate a downturn multiplier, and export the final scored loan book with a full audit trail.
Full-stack demo that classifies a loan portfolio into IFRS 9 Stage 1 / 2 / 3 and calculates Expected Credit Loss at loan and portfolio level, with configurable SICR thresholds and CSV upload.
Calculates Basel IRB risk-weighted assets, Pillar 1 capital requirements and regulatory expected loss across retail and corporate/SME exposure classes, with asset correlation, maturity adjustment and SME firm-size support.
Calculates single-asset Value at Risk and Expected Shortfall using Historical Simulation, Variance-Covariance and Monte Carlo methods side-by-side, with backtesting (Kupiec POF test) and reference implementations in five languages.
Options Pricer & Greeks Dashboard
Prices European, American and barrier options using Black-Scholes, Monte Carlo and binomial tree methods, with full Greeks, implied volatility smile, convergence analysis and P&L scenario heatmap.
All projects are personal and use synthetic data, simulated examples or publicly available sources. Nothing here reflects the work, data or intellectual property of any employer or client.