About the job
We’re starting to see the incredible potential of multimodal foundation and large language models, and many applications in the computer vision and machine learning domain that previously appeared infeasible are now within reach. We are looking for a highly motivated and skilled Machine Learning Integration Engineer to join our team in the Video Computer Vision group and help us ship cutting edge computer vision technology on Apple devices. The Video Computer Vision org has pioneered features such as FaceID, FaceKit, and Gaze and Hand gesture control which have changed the way millions of users interact with their devices. We balance research and product requirements to deliver Apple quality, pioneering experiences, innovating through the full stack, and partnering with HW, SW and AI teams to shape Apple's products and bring our vision to life.
Responsibilities
Convert and integrate foundation models for on-device deployment.
Evaluate power and performance of models running on Apple devices.
Optimize on-device inference latency and efficiency of CV/ML models.
Qualifications
Minimum
BS and a minimum of 10 years relevant industry experience.
Strong knowledge of model compression techniques such as pruning, distillation, quantization and weight clustering.
Solid understanding of operating system and extensive programming experience in Python and C++.
Experience working with PyTorch.
Experience with machine learning model development lifecycle, including data preprocessing, model training, evaluation, and deployment.
Foundational understanding of machine learning: MultiModal LLMs and integration of ML components into production systems.
Preferred
Experience with CoreFoundation, RealityKit and CoreML frameworks.
Fundamental knowledge of real-time video pipelines, image transformations, and rendering loops.
Programming experience with Swift.