Economist, quant, and data scientist
I use econometrics and machine learning to measure how economic shocks move shipping, commodity, and financial markets, and to forecast where they go next.
I’m a senior economic analyst and financial engineer with more than 18 years of experience combining rigorous economic analysis with quantitative modeling. I build models that connect the wider economy to market outcomes, and I explain what they mean to the people who have to make decisions with them.
My work ranges from predictive models and financial risk management to the data infrastructure that decisions rely on. I’m also a published technical author, and I spent 16 years teaching economics at university.
– present
National Navigation Company, Cairo
I research and forecast the maritime and shipping industry, measuring how global trade dynamics and economic shocks affect the sector. I build econometric, machine learning, and deep learning models that link macroeconomic variables to sector performance and inform the company’s strategic planning. I work in Python with NumPy, pandas, scikit-learn, TensorFlow, and PyTorch.
I also use financial modeling, scenario analysis, and risk management to improve operational efficiency, assess market opportunities, and limit exposure to market volatility. I build and deploy production Power BI dashboards with scheduled refresh, and I maintain the economic databases behind our modeling and reporting. I write technical reports and policy briefs in Arabic and English for executives and government bodies, including the Ministry of Transport and port authorities.
–
Udacity, remote
I led live virtual sessions for Udacity’s Data Analysis, Data Science, and AI Trading Strategies Nanodegree programs, teaching Python, SQL, statistical modeling, and machine learning for finance. Learners rated my sessions 4.9 out of 5 on average. I also reviewed student projects in data science and Python with detailed feedback, contributed to curriculum and rubric design, and wrote technical articles for the Udacity blog. Through the Digital Egypt Cubs Initiative (DECI), I taught the foundations of data science and AI to students aged 12 to 17.
–
Alexandria University, Alexandria
For 16 years, I taught econometrics, mathematical economics, microeconomics, macroeconomics, and international economics to classes of up to 150 students. I supervised and assessed student research projects, helping students develop analytical rigor and clear written communication. I also advised students, helped administer and run examinations, and co-founded the Economic Society to deepen students’ engagement with economics.
An XGBoost forecasting system for Panamax dry bulk freight rates, built on 245 features engineered from more than 700 market indicators. I also found a systematic timing bias in the forecasts and corrected it with phase-shift calibration.
| Measure | Result |
|---|---|
| Reduction in RMSE after phase-shift calibration | 34% |
| Directional accuracy | 77% |
| Sharpe ratio in commercial validation | 2.0–3.1 |
| Engineered features, from 700+ market indicators | 245 |
A Python tool for bilateral trade analysis that combines harmonized BACI data (1995–2024) with the live UN COMTRADE API (1962–present). It has an interactive command-line interface, fuzzy country matching, multi-level HS code support, professional visualizations, and comprehensive data exports for researchers, policy analysts, and trade professionals.
Single-output and multi-output LSTM models that predict 25-day returns for five ETFs across equities, bonds, gold, and oil. The predictions drive long-short strategies that outperformed buy-and-hold benchmarks on risk-adjusted returns in backtests, and the project also covers tactical asset allocation.
Random forest, MLP, and TabNet models that forecast sales volume from product positioning, pricing, and promotion data, with statistical testing (ANOVA, Shapiro–Wilk) and feature engineering to identify the main sales drivers.
A minimal inventory management system written in Go with Gin and GORM, featuring CRUD endpoints with UUIDs, rate limiting, and PostgreSQL integration, plus pagination, filtering, and sorting.
An analysis and modeling project that predicts Bitcoin prices with multi-layer perceptron (MLP) and convolutional neural network (CNN) models, including data preprocessing.
An investigation of how usage differs between customers and subscribers of Bay Wheels, the San Francisco Bay Area bike-share system, and of the main factors that affect trip duration.
I’ve written practical guides on data science, machine learning, and Python for the Udacity blog.
–
WorldQuant University, online
Graduated with a 94% cumulative average and a grade of Excellent.
Capstone: Advanced Freight Rate Forecasting for Panamax Dry Bulk Carriers using Machine Learning.
Coursework in financial econometrics, risk management, machine learning and deep learning for finance, stochastic modeling, derivative pricing, and portfolio management.
–
Aarhus University, Denmark
Studied on a Danish State Scholarship.
–
Alexandria University
Completed with a grade of Good (78.66%).
–
Alexandria University
Graduated with honors, grade Very Good (88.5%), top of class.
Microeconomics, macroeconomics, and international and maritime economics. Econometric modeling, time series analysis, statistical inference, hypothesis testing, and regression analysis.
Financial modeling, portfolio and risk management, scenario analysis and stress testing, derivative pricing, and stochastic modeling.
Python (NumPy, pandas, scikit-learn, Statsmodels, XGBoost, TensorFlow, PyTorch), SQL, advanced Excel, Git and GitHub, and some R. Visualization with Matplotlib, Seaborn, and Power BI.
Published technical author and 16 years of university teaching. Technical writing in Arabic and English. Arabic (native), English (full professional), Danish (elementary).
28 courses and 6 tracks covering Python, pandas, statistics, regression, and data visualization.
Email is the quickest way to reach me. I’m based in Cairo and work in English and Arabic.