Shahzeb K.

I build and audit forecasting and trading models that hold up out of sample, and I explain the results in plain financial terms. Recent work: for a Swedish asset manager I built a five-year, minute-level research pipeline across seven European index markets (about 8.4 million bars), trained XGBoost and Temporal Fusion Transformer models with walk-forward validation, and audited every backtest for look-ahead bias. The client has hired me three times: "a true expert in financial analysis and also strong in coding with Python." What I can do for you • Machine-learning forecasting and trading research: Python, XGBoost, TFT, LSTM, SHAP • Backtest and model audits: leakage checks, walk-forward and purged cross-validation, deflated Sharpe ratio, golden-master tests • Econometrics and risk: panel data, difference-in-differences, GARCH, VAR and connectedness, event studies • Financial modelling: cash-flow forecasts, valuation and scenario analysis in Excel or Python Background: PhD in Finance; postdoctoral research fellow at Ege University, Türkiye; finance lecturer since 2018; peer-reviewed work in Finance Research Letters, Financial Innovation and Mineral Economics. How I work: clear milestones, reproducible code, short written reports, and an honest answer when a strategy or model does not work. Send me your data or model and the decision it supports, and I will reply with a scoped plan and a fixed quote.