Job Description
Video is at the core of nearly all Apple products, and as a research engineer on our team, you will build the infrastructure, pipelines, systems, and models that power the next generation of video technology. We are looking for a highly ambitious individual, who will flourish working on technically meaningful problems across the full ML lifecycle — including training infrastructure, performance optimization, data, and production deployment. Your work will redefine the video experience for billions of users.
In this role you will work together with colleagues to design, scale, and harden the systems that bring ML-based video approaches into current and future Apple products. This position requires a highly self-directed individual, who is comfortable working at the intersection of ML, systems engineering, and performance optimization. Strong engineering and analytical skills will be critical towards solving challenging problems across efficiency optimization, training, data, and deployment.
Preferred Qualifications
MS specializing in ML systems, performance engineering, or a related area
Experience implementing custom ops in CUDA or low-level GPU kernel optimization
Research experience in ML model design
Experience with visual data (e.g. images/videos/3DGS) and related vision algorithms
Experience with distributed training, large-scale data pipelines, inference serving
Excellent written and oral communication skills
Minimum Qualifications
BS and 10+ years of hands-on industry experience building production ML systems, with a proven track record of design/implementation leadership
Deep familiarity with in of ML-centric flows and best practices: training procedures, dataloaders, profiling and debugging methodology, data handling, model development, model deployment
Proficiency in Python, C++, PyTorch
Familiarity with computer architecture principles
Proficiency in AI tooling such as Claude and excellent high-level design skills — allowing building robust, modular, clean, and well-tested code
