Data Scientist ML, Forecasting, Fraud Detection Expert

ABDELKABIR SAHNOUN

DATA SCIENTIST

Raleigh, NC **606 • 1-984-***-**** • *******.**********@*****.***

linkedin • github • kaggle

Data Scientist with 5 years of experience in ML, statistical modeling, and cloud deployment. Experienced in fraud detection, churn prediction, time-series forecasting, and customer analytics. Skilled in Python, R, SQL, and AWS, with a proven record of delivering measurable business results through data-driven insights and end-to-end ML solutions. CORE SKILLS & INTERESTS

• Programming: Python, C/C++, Java, SQL, R

• Machine Learning: Scikit-Learn, XGBoost, Deep Learning, TensorFlow, Keras, Prophet, ARIMA

• Data Visualization: Tableau, Power BI, Matplotlib, Seaborn, Excel

• Cloud & Tools: AWS, Git/GitHub, Flask API, Jupyter, AI, Statistical Modeling, SHAP

• Specialties: Supervised learning, Unsupervised learning, Model Selection, Predictive Analytics, Fraud Detection, Time-Series Forecasting, Churn Prediction, Customer Segmentation, A/B Testing, Evaluation EXPERIENCE

MOVE Fellow – AI Expert 10/2025 – Present

Handshake AI – Project Orion (Remote)

• Evaluated and corrected rogue behaviors in AI research agents to ensure scientific validity and ethical compliance.

• Developed Bug Fix Plans and implemented corrected ML code improving model reliability and reproducibility.

• Contributed to annotation frameworks and alignment rubrics for large-scale AI reasoning evaluation.

• Collaborated with researchers to enhance prompt engineering and Responsible AI practices. Data Scientist 08/2021 – 08/2025

InnovatiCS, Alpharetta, Georgia, USA

• Fraud Detection in E-Commerce - Built XGBoost-based fraud detection model, achieved Precision = 98%, deployed via Flask on AWS EC2 with SHAP interpretability. GitHub Repo Kaggle Notebook

• Customer Segmentation & Behavior Prediction - Applied K-Means clustering and PCA to segment customers; built predictive models and dashboards to support retention strategies. GitHub Repo Kaggle Notebook

• Credit Card Churn Prediction - Developed Random Forest and XGBoost models for churn prediction; integrated SHAP values for interpretability and deployed API for real-time scoring. GitHub Repo Kaggle Notebook

• Amazon Stock Price Forecasting - Designed ARIMA, Prophet, LSTM, and XGBoost time-series forecasting models; achieved less than 1.1% prediction error rate; deployed best model in web application. GitHub Repo Kaggle Notebook

• E-Commerce A/B Testing Framework - Created statistical experimentation pipeline to analyze conversion lift and statistical significance for checkout optimization. GitHub Repo Kaggle Notebook ADDITIONAL WORK EXPERIENCE

Research associate, Networking and Computer Science 02/2018 - 02/2020 San Diego State University, San Diego, CA

Designed and implemented an adaptive low overhead Link State routing scheme for Unmanned aerial vehicles UAVs. EDUCATION

PhD, Networking and Computer Science – Mohammed V University, Rabat, MA – 09/2019 Research associate, Networking and Computer Science – San Diego State University, San Diego, CA – 02/2020 Master, Computer science and Telecommunication – Mohammed V University, Rabat, MA – 07/2013 Bachelor, Mathematics and Computer science – Mohammed V University, RABAT, MA – 07/2011 PROFESSIONAL DEVELOPMENT

• Data Science & Artificial Intelligence: Certification of training program in DS & AI completion, by Innovatics.

• Google Data Analytics Professional Certificate: Certification course by Google, offered through Coursera.

• Machine Learning: Certification course by Stanford University, offered through Coursera.

Contact this candidate