About the Role

At Aevum Zenth, computer vision isn't just a tool—it's the foundation of our autonomous future. In this role, you'll lead the development of multimodal perception pipelines that power everything from orbital satellite navigation to sub-millimeter surgical robotics. You'll collaborate with world-class researchers, engineers, and domain experts across our Aerospace, Healthcare, and Robotics divisions to solve problems that have never been solved at scale.

This is a research-forward engineering role. You'll publish, prototype, and productionize. You'll have access to dedicated GPU clusters, petabyte-scale proprietary datasets, and direct pathways to commercialize your breakthroughs across our global portfolio.

Key Responsibilities

• Design and implement state-of-the-art computer vision architectures for 2D/3D object detection, semantic segmentation, optical flow, and visual SLAM.
• Develop multimodal fusion pipelines combining RGB, LiDAR, radar, and thermal imagery for edge-deployed autonomous systems.
• Optimize models for latency-constrained environments (Jetson, custom ASICs, FPGA) using TensorRT, ONNX, and custom CUDA kernels.
• Lead data strategy initiatives: synthetic data generation, active learning frameworks, and automated annotation pipelines.
• Publish breakthrough research in top-tier venues (CVPR, ICCV, ECCV, NeurIPS) and patent core perception IP.
• Mentor junior scientists and engineers, fostering a culture of rigorous experimentation and production-ready engineering.

What We're Looking For

• Ph.D. or M.S. in Computer Science, Electrical Engineering, Robotics, or related field.
• 5+ years of experience in computer vision research or applied ML engineering.
• Deep expertise in PyTorch/TensorFlow, OpenCV, and modern training frameworks (DeepSpeed, Megatron, Lightning).
• Proven experience deploying CV models to production at scale (Kubernetes, Triton, TFX, or similar).
• Strong mathematical foundation in linear algebra, probability, optimization, and differential geometry.
• Track record of publications, open-source contributions, or patented vision technology.

Nice-to-Haves

• Experience with neuro-symbolic AI or vision-language models (VLMs).
• Background in aerospace guidance/navigation or medical imaging regulation (FDA/CE).
• Contributions to open-source vision libraries (OpenMMLab, Detectron2, Hugging Face).
• Experience leading cross-divisional technical initiatives.

Why Aevum Zenth?