Generative AI Applied Scientist, SIML - ISE

About the role

Apple's System Intelligence and Machine Learning (SIML) team is seeking a senior Generative AI expert to pioneer the next generation of human-centric device interaction and multimodal scene understanding. You will be at the core of efforts to develop multimodal LLMs that can perceive and understand complex scenes and nuanced human interactions, behaviors, and preferences. This is a unique opportunity to join a leading applied research group known for its foundational contributions to Apple Intelligence, where you will focus on the end-to-end lifecycle of generative models—from novel architecture design and large-scale training to final deployment.

Responsibilities

• Design, train, and deploy multimodal GenAI technology
• Learn quickly and implement and demonstrate new user experiences using large foundation models
• Build novel and innovative technology, forge collaborations with cross-functional partners, and adapt and iterate solutions in a dynamic environment
• Advance human interaction and scene understanding modeling across various fronts, from a system level to a core ML algorithm level
• Address challenges including understanding user behavior and preferences from interactions with the device and the environment
• Retrieve useful and nuanced information based on past interactions
• Handle deeply interleaved streaming inputs
• Reason over varying temporal contexts
• Develop memory systems to enable long-term adaptation
• Generate semantically rich internal representations to enable open-ended downstream tasks
• Deliver ML models and solutions that can readily be adopted in production pipelines, such as APIs for production-ready ML models and algorithms well-integrated into training infrastructure

Minimum qualifications

• PhD or Masters Degree in Computer Science, Engineering, or a related field with a focus on machine learning; or equivalent experience
• Strong research skills with first author publications in top tier ML conferences
• Expert-level knowledge of SOTA in large auto-regressive transformer models, multi-modal encoders, and representation learning
• Experience with multimodal large language models (LLMs)
• Strong programming skills in Python, maintaining ML code bases grounded in software engineering principles

Preferred qualifications

• Proven track record of deploying innovative ML technologies in production
• Familiarity with developing ML for resource-constrained devices
• Experience working with large cross-functional and diverse teams