Data & Quantitative Finance Analyst | CPA Finalist · Nairobi, Kenya

I build quantitative models that turn financial data into decisions — from LightGBM fraud classifiers with SHAP attribution to CFA-level portfolio analytics and DCF valuation workbooks.

I work across the full analytics stack: Python-based ML models (LightGBM, scikit-learn) and CFA-level financial modelling (VaR, CAPM, CVaR, Sharpe/Sortino ratios) to production-grade Excel dashboards and DCF valuation workbooks.

Analytics & Data Science

ML & Finance

Credit card fraud classifier on the Kaggle Playground S3E4 dataset (219K transactions). LightGBM at 99.11% precision / 96.88% recall on a 132K time-based holdout. Implements balance-discrepancy features, calibrated probabilities, and SHAP explainability.

Python · LightGBM · scikit-learn · Jupyter · Imbalanced Learning · SHAP

Professional-grade portfolio management system with live market data, CFA-level risk decomposition (VaR, CVaR, CAPM, Jensen's Alpha), Black-Litterman optimisation, automated rebalancing, and a 23-test validation suite.

Excel 365 · Financial Modelling · Risk Analytics · Portfolio Management

Company financial analyses built from primary-source filings: three-statement models, DCF valuations, and scenario toggles. Each workbook has a validation tab that ties every historical line to its source.

Excel · DCF Valuation · Fundamental Analysis · Equity Research

• 🗄️ SQL-based financial data pipelines for portfolio analytics
• 📊 Power BI dashboards for financial KPI reporting
• 🧠 Expanding ML work into credit risk and time-series forecasting

Open to: Data Analyst · Finance Analyst · Quantitative Analyst · Financial Data Scientist · FinTech Analytics roles