I am Eslam Sharaawy,
a data scientist
& ML developer
based in Germany.

Working on multiple AI research projects across scientific machine learning, probabilistic modeling, generative AI, and foundation models. Research includes predicting battery State of Health (SOH) and Remaining Useful Life (RUL) using probabilistic and generative models, with published work in this area. Additional projects focus on identifying inflow locations in Lower Saxony by modeling hidden subsurface network structures from precipitation data. Also contributing to the development of foundation models for scientific applications, including physics-informed neural networks (PINNs), operator embeddings, and related AI methods.

Developed a damage cost prediction model combining ResNet-based visual features with vehicle data, using CatBoost to cut RMSE by 43%. Accelerated deep learning inference by 35% via hyperparameter tuning and data augmentation. Proposed AI-driven process improvements and, for a master’s thesis, built a few-shot learning method for detecting rare vehicle damages.

Tested and validated new systems to identify and track errors in close collaboration with the development team. Automated reports using VBA, resulting in a 30% reduction in manual processing time. Supported the team in daily tasks to enhance overall efficiency.

Simulated a component of an injection molding machine by varying key process parameters and created a predictive model based on the simulation data to forecast outcomes without the need for repeated simulations. Utilized this data-driven approach to support the development of a digital twin, enabling real-time performance monitoring and predictive maintenance.

Delivered technical support for servers, Exchange migrations, and client connectivity, escalating issues as needed to ensure stable IT operations. Maintained and secured infrastructure via backups, disaster recovery, system hardening, and compliance with security policies. Handled daily IT tasks, including documentation, storage planning, workstation/server upkeep, and software/network performance optimization.

Managed a major project in Egypt’s new administrative capital on behalf of Siemens, ensuring quality control and implementation of mechanical designs in accordance with internal Siemens standards, and provided regular progress reports to the client.

Achieved a final grade of 2.1 in the German grading system, with strong expertise in mathematics, programming, statistics, simulations, and physics applied to real-world problems. Specialized in artificial intelligence and data science, including data analysis, regression, classification, data modeling, classical machine learning methods such as XGBoost and Random Forest, and advanced deep learning methods such as CNNs, attention mechanisms, and transformer models. Gained practical experience in generative AI, computer vision, large language models, classical AI, and physics-informed neural networks. Completed a student project on semantic segmentation of infrastructure cracks using machine learning and deep learning. For the master’s thesis in collaboration with Volkswagen Financial Services, developed a few-shot computer vision pipeline that reduced annotation costs by 35% and enabled rapid detection of rare defects.

Holds a Bachelor’s degree in Mechanical Power Engineering from Helwan University, Faculty of Engineering in El Matareya, completed in May 2019. Graduated with an overall grade of 86.6% and the distinction “Excellent,” including an “Excellent” rating for the final project. The cumulative GPA of 80.44% earned the classification “Very Good with Honors.” The program, accredited by Egypt’s National Authority for Quality Assurance of Education, provided a strong foundation in core areas such as mathematics, physics, mechanical design, statistics, programming, and power systems, with consistently high performance throughout the five years of study.