AI Research Scientist (Multimodal Foundation Models)

Quick Facts

• Role: AI Research Scientist focused on multimodal foundation models

• Focus areas: architecture, pre-training, scaling, distillation, and evaluation for on-device intelligence

Description

You will investigate how multimodal foundation models should be designed, trained, and distilled, with a strong emphasis on transferring capabilities from large models into smaller, efficient ones. Your work includes developing and evaluating new architectures, pre-training objectives, data strategies, optimization methods, and teacher–student learning techniques while incorporating compute, memory, latency, energy, and privacy constraints throughout the research and evaluation process.

Responsibilities

• Develop and evaluate model architectures for multimodal understanding and efficient on-device inference

• Design and test pre-training objectives, data mixtures, tokenization/curricula, and optimization methods

• Advance distillation approaches that preserve reasoning, multimodal understanding, and instruction following under constrained model capacity

• Run large-scale experiments across model research, data, systems, and hardware boundaries

• Translate successful research into foundation-model technologies; may publish and open-source selected models and artifacts

Requirements

• Hands-on experience designing, implementing, and running large-scale pre-training experiments for large language models

• Experience with LLM pre-training topics: architecture, training objectives, data mixtures, tokenization, curricula, scaling, optimization

• Strong proficiency with PyTorch or JAX and distributed training systems

• Experience evaluating pre-trained models across language understanding, reasoning, instruction following, or multimodal capabilities

• Strong understanding of transformer-based architectures and efficient/scalable foundation-model training approaches

• Master’s degree (or equivalent practical experience) in machine learning, computer science, or related technical field

• Minimum 6 years of relevant experience

Benefits

• Opportunity to shape new research directions and run ambitious experiments at scale

• Chance to co-develop frontier and efficient multimodal models optimized for Apple silicon and on-device intelligence

• Potential path to publications and open-sourcing research artifacts, evaluations, or tools