I'm an ML/MLOps Engineer based in Germany, currently working at APTIV on the Computer Vision Interior Sensing team. I specialize in building scalable machine learning infrastructure, from data pipelines to distributed training systems.

My journey spans 8+ years across automotive software, data science, and machine learning. I started as a software engineer at Valeo building AUTOSAR-compliant ECU components, then transitioned into ML through my Master's at TU Munich, where I focused on computer vision and deep learning.

What I Do

  • MLOps & Infrastructure: Design and manage hybrid Kubernetes clusters (on-prem + cloud), build Airflow pipelines for automated training workflows, and implement GitOps with ArgoCD
  • Computer Vision: Work on perception systems for automotive applications, including depth estimation, segmentation, and object tracking using Graph Neural Networks
  • Cloud & Distributed Systems: Manage large-scale GPU clusters on AWS/Azure, optimize distributed training on SLURM, and build cost-efficient LLM/VLM inference platforms
  • Data Engineering: Handle terabyte-scale CV datasets - acquisition, annotation, versioning with DVC, and validation pipelines

Background

I hold an M.Sc. in Informatics from TU Munich (focus: Machine Learning & Computer Vision) and a B.Sc. in Mechatronics Engineering from the German University in Cairo. My research has been published in venues covering biomedical sensing and robotics.

Beyond Work

I contribute to open source (BlenderProc for synthetic data generation) and enjoy exploring the intersection of embedded systems and modern ML infrastructure - like my recent work on Kubernetes extensions for Hardware-in-the-Loop testing.

View my full resume →