Yusuf Sami
ABOUT / YUSUF SAMI
From operational data
to better decisions.
I study Industrial Engineering at MEF University while completing a minor in Data Science and Artificial Intelligence. During my Erasmus exchange at Eindhoven University of Technology, I took MSc-level coursework in transport, mobility, and operations management. A recent role at Volt Lines gave me first-hand exposure to transportation planning and real-world operating constraints—and confirmed that I want to move closer to technical work in data science, predictive modeling, and optimization.
THE THROUGH-LINE
I work across the full decision pipeline.
My work starts with operational questions: what is happening in the system, which patterns are reliable, and what action should follow? Working close to day-to-day transportation operations showed me how planning decisions are shaped by demand, vehicle availability, driver assignments, time windows, and regional constraints. It also clarified what I want from my next role: deeper ownership of analysis, modeling, experimentation, and optimization.
In my projects, I use Python, statistical analysis, machine learning, simulation, and mathematical optimization as connected tools. A low prediction error is useful only if the resulting maintenance plan improves. An optimal route matters only if it respects capacity, time windows, driver limits, and uncertainty.
This is the kind of Data Science I want to practice: technically grounded, explicit about assumptions, and close enough to the operation to change a real decision.
- 01
Read the system
Structure, validate, and explore data with SQL, Python, and domain context.
- 02
Measure uncertainty
Use statistics, bootstrap inference, simulation, and stress tests to separate signal from noise.
- 03
Build the model
Choose predictive or optimization methods around the decision—not around a fashionable algorithm.
- 04
Validate the consequence
Check feasibility, cost, capacity, service quality, and how the recommendation behaves in practice.
CAPABILITY MAP
Problems I like working on.
Operational data with context
SQL, Python, data validation, pattern analysis, and decision-oriented reporting.
Predictions with consequences
Feature engineering, XGBoost, PyTorch, calibration, and downstream evaluation.
Action under constraints
Routing, assignment, location, scheduling, capacity, time windows, and uncertainty.
Results that survive error
Bootstrap inference, sensitivity analysis, Monte Carlo simulation, and stress tests.
Flow, waiting, and capacity
Queueing models and discrete-event simulation for operational systems.
Models grounded in movement
Passenger demand, vehicle utilization, routing policies, and transport operations.