Experience
Building Machine Learning Systems for Complex, High-Stakes Environments
Six roles across three countries, each one sharpening the same question: how do risks interact, and how do we predict them?
Postdoctoral Fellow
McMaster University · Faculty of Engineering
This is where the threads of my PhD come together. I lead research on predictive intelligence for risk and resilience in complex systems, combining machine learning with network analysis, optimization and digital twins to anticipate how failures spread, from wildfire and flood hazards to cascading risks across interdependent infrastructure. Increasingly, that work extends to language models that can reason over these risk networks.
Research Engineer
McMaster University · DeGroote School of Business, Smart Freight Centre
Alongside my PhD, I took the same risk-prediction thinking into the air. At the Smart Freight Centre, I build real-time models that predict accident risk for drones carrying hazardous materials, work that now supports larger research programs in AI-driven logistics and Advanced Air Mobility.
Deputy Project Manager
WSP Canada Inc.
An industry placement let me see risk from the other side of the desk. Leading infrastructure project teams at WSP showed me how data-driven risk management changes outcomes on real budgets and real schedules, a perspective that still shapes the systems I build.
Doctoral Researcher
McMaster University · Faculty of Engineering
This is where the core idea took shape: risks in complex systems rarely act alone; they interact and cascade. I built machine learning and graph-based models to predict those interactions, moved them into production pipelines on Google Cloud, and shared the results through peer-reviewed research.
Researcher & Teaching Assistant
The German University in Cairo (GUC)
After graduating, I stayed at GUC to teach, and started asking whether machine learning could change how we manage infrastructure. That question became the foundation of my PhD.
Bachelor Thesis Researcher
University of Stuttgart · ILEK Institute, Germany
A six-month research placement in Stuttgart was my first serious encounter with computational modeling. Building a C#/.NET simulation tool to analyze how complex systems respond shifted the direction of everything that followed.
Leadership & Service
Building the Research Community
- 2026Local Organization CoordinatorFuture of Our Planet Conference (FOP2026)McMaster University
- 2025Operations CoordinatorNuclear Renaissance 2.0 ConferenceMcMaster University
Technical Deep Dive
Tools and Methods Across These Roles
Machine Learning & Deep Learning
Designing and training predictive models
Frameworks
- Python
- PyTorch
- TensorFlow / Keras
- PyTorch Lightning
- scikit-learn
- XGBoost
- LightGBM
- CatBoost
Architectures
- CNN
- LSTM
- ConvLSTM
- Transformers
Results & Metrics
What These Roles Delivered
Technical Impact
- Better Wildfire Prediction
- Improvement over baseline models in predicting wildfire hazard from geospatial time-series data.
- Faster Deployment
- Through reproducible, production-grade TFX pipelines on Google Cloud.
- Training Efficiency Gain
- Through optimization-based hyperparameter tuning and parallelized search strategies.
Research & Community Impact
- CAD in Research Proposals
- Contributed to funding proposals for AI-driven logistics and freight research.
- Peer-Reviewed Journal Papers
- Plus 2 conference papers, spanning machine learning, network analysis and infrastructure risk.
- Package Downloads
- Three open-source ML tools adopted by practitioners worldwide.
Education
Academic Background
Engineering Ph.D., Machine Learning for Systemic Risk Management of Complex Systems
Vanier Scholar · McMaster University, Canada
Engineering B.Sc., Class Valedictorian, Highest Honors
The German University in Cairo (GUC), Egypt
Engineering Bachelor Thesis
University of Stuttgart, Germany
Let's Build Something That Matters
If you have a decision that depends on predicting risk in a complex system, I'd like to hear about it.