Ahmad Abdel-Qader

Leads the end-to-end delivery of production LLM systems and 2D/3D medical-imaging models, spanning model development, scalable inference, regulatory validation, monitoring, and clinical operations.

• Built and productionized LLM services and 2D/3D medical-imaging inference systems that process thousands of requests every day.
• Reduced production ML inference time by 7× through end-to-end optimization of model execution and serving architecture.
• Leads a cross-functional team of seven ML engineers, DevOps, and full-stack developers delivering production clinical AI products.
• Architected scalable Kubernetes model-serving infrastructure for high-availability, low-latency AI deployment.
• Shipped CT nodule and chest X-ray abnormality detection to clinical use, reaching 96% sensitivity under rigorous statistical validation.
• Orchestrated large-scale model deployment across multi-cloud AWS and GCP environments.
• Set the testing, monitoring, and compliance standards for AI systems under regulatory requirements and Responsible AI principles.
• Improved classification accuracy 10% by integrating state-of-the-art backbones into the existing production codebase.

• Python
• PyTorch
• LLM Systems
• 2D/3D Medical Imaging
• Model Serving
• MLOps
• Kubernetes
• AWS
• GCP
• Computer Vision

Conducted graduate research at the intersection of information theory and machine learning, developing and evaluating learning-based methods for wireless communication systems.

• Applied machine learning and statistical modeling to optimize communication systems and evaluate performance.
• Authored peer-reviewed research translating theoretical methods into documented system improvements.
• Developed scalable simulation and analysis workflows for real-world communication-system design.

• Python
• MATLAB
• Machine Learning
• Information Theory
• Simulation

Built and led an ASIC functional verification team, and implemented full-chip digital and mixed-signal designs from RTL through GDSII.

• Built and led a functional verification team from scratch, mentoring engineers and accelerating project delivery.
• Developed complex testing frameworks to validate system performance against standard protocols.
• Implemented full-chip digital and mixed-signal designs from RTL to GDSII in deep sub-micron technologies.

• RTL Design
• Functional Verification
• RTL-to-GDSII

Led GPS-denied indoor positioning research combining human activity recognition with pedestrian dead reckoning, and built the IMU-based navigation stack it ran on.

• Led development of deep learning activity recognition using LSTM architectures for complex sequence modeling.
• Reached 95%+ classification accuracy while evaluating model robustness against real-world sensor conditions.
• Designed efficient algorithms for mobile platforms, handling sensor noise through advanced filtering and optimization.
• Built an IMU-based navigation system and contributed to smartwatch prototype hardware design.

• Python
• LSTMs
• Sensor Fusion
• Embedded ML