ML Systems Engineer: End-to-End Training & On-Device

Apple Inc.; Pittsburgh (Allegheny County); On-site; USD 150,000 - 220,000; Full time

Job summary

Apple Inc. in Pittsburgh is seeking a Machine Learning Engineer for the Machine Intelligence Neural Design (MIND) team to advance cutting-edge ML features. You will work on Foundation Models, Perception, and end-to-end ML development from data curation to on-device optimizations.

The role emphasizes shipping ML-based features, collaboration with ML researchers and SW/FW engineers, and building real-time demos and visualizations of sensing data streams and model predictions.

Qualifications

• PhD in computer science, computer engineering, or relevant fields; BS/MS with 3–5 years ML eng exp also qualifies.
• Strong foundation in machine learning, especially LLM and multimodal foundation models.
• Experience in building model training/eval pipelines in Python/PyTorch.
• Experience in prototyping and developing software applications (preferably in Swift).
• Experience with sensors and sensing systems.
• Strong communication and presentation skills.
• Ability to work in a collaborative environment.

Responsibilities

• Build ML models and end-to-end training and evaluation pipeline.
• Develop ML production software.
• Ability to build end-to-end demos for ML solutions.
• Represent the team in Cross-functional discussions.

Skills

LLM/multimodal models

Training pipelines

Collaboration

Communication

Sensing systems

Swift (prototyping)

Education

PhD in CS/CE or related field

BS/MS + 3–5 years ML eng experience

Tools

Python

PyTorch

Swift

JAX

TensorFlow

Job description

Apple Inc. in Pittsburgh is seeking a Machine Learning Engineer for the Machine Intelligence Neural Design (MIND) team to advance cutting-edge ML features. You will work on Foundation Models, Perception, and end-to-end ML development from data curation to on-device optimizations.

The role emphasizes shipping ML-based features, collaboration with ML researchers and SW/FW engineers, and building real-time demos and visualizations of sensing data streams and model predictions.