Sr ML Research Engineer, Post-Training for Personalization

Apple • Cupertino, CA

14h

About The Position

Apple’s Intelligent System Experience team develops core modeling technologies that bring cutting-edge research into user experiences across Apple Intelligence initiatives, including Personalization, Agentic Search, Generative Memories, and Visual Generation.

The role focuses on researching and developing novel LLM post-training algorithms for personalized intelligence. The engineer will advance methods that help models understand individual user needs, adapt to personal context, and provide more relevant, helpful, personalized experiences. Work spans large language models, multimodal reasoning, personalization, and agentic systems, translating research into scalable, high-quality experiences.

The position involves developing algorithms for prompt optimization and post-training, applying techniques such as reinforcement learning, multimodal reasoning, context compaction, efficiency methods, and memory systems. It also offers opportunities to partner with system engineers to prototype and incorporate new algorithmic innovations into emerging agentic experiences. Familiarity with large-scale modeling and data infrastructure is a plus.

Requirements

• PhD in Electrical Engineering, Computer Science, or a related field (such as mathematics, physics, or computer engineering), focused on computer vision and/or machine learning, or comparable professional experience.
• Strong machine-learning and generative-modeling fundamentals.
• Hands-on experience in one or more of: pre-training or post-training large language models, and reinforcement learning.
• Familiarity with distributed training.
• Proficiency with machine-learning toolkits such as PyTorch.
• Research contributions demonstrated through publications in top-tier conferences or open-source contributions to algorithms.
• Fluency in algorithm development, including prompt optimization and post-training.
• Strong expertise in relevant techniques such as reinforcement learning, multimodal reasoning, context compaction, efficiency, and memory systems.

Nice To Haves

• Experience building and deploying Multimodal LLMs.
• Experience building and deploying agentic systems and LLMs with tool-use and memory capabilities.
• Ability to interface with large-scale modeling and data infrastructure.

Responsibilities

• Drive research in post-training methodologies for personalized intelligence.
• Develop algorithms that help models understand individual user needs and adapt to personal context.
• Translate emerging research into scalable, high-quality experiences involving LLMs, multimodal reasoning, personalization, and agentic systems.
• Prototype and incorporate algorithmic innovations into emerging agentic experiences in partnership with system engineers.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: Ph.D. or professional degree