Senior ML Platform Architect for GenAI & Data Pipelines

Apple; Cupertino (CA); On-site; USD 250,000 - 320,000; Full time

Job summary

Apple is seeking a Staff ML Infrastructure Engineer to own the architecture of the platform powering Apple's large-scale model builds. You will define ingestion, versioning, lineage, and governance for petabyte-scale data, and design high-throughput data delivery to the GPU/TPU fleets.

You will mentor senior engineers, lead design reviews, and drive the platform roadmap for next-generation GenAI workloads, embeddings, and retrieval-augmented systems.

Qualifications

• 10+ years in ML infrastructure or distributed data systems.
• Experience building large-scale data or ML platforms in production.
• Deep systems engineering with Python and a systems language (Rust preferred).
• Familiarity with Parquet, Iceberg, Delta, Lance formats.
• Strong collaboration and mentorship.

Responsibilities

• Own the platform architecture behind large model builds across ingestion, versioning, lineage, and governance at petabyte scale.
• Define data access and loading architecture to keep training compute-bound, not IO-bound.
• Lead design reviews, mentor engineers, and align multiple teams on technical direction.
• Drive platform roadmap for foundation models, multimodal data, and retrieval-augmented systems.

Skills

Python

Systems design

Performance engineering

Collaboration

Mentoring

Education

B.S./M.S./Ph.D. in Computer Science or Computer Engineering

Tools

Rust

C++

Go

Parquet

Iceberg

Delta

Lance

Arrow

Docker

Kubernetes

Job description

Apple is seeking a Staff ML Infrastructure Engineer to own the architecture of the platform powering Apple's large-scale model builds. You will define ingestion, versioning, lineage, and governance for petabyte-scale data, and design high-throughput data delivery to the GPU/TPU fleets.

You will mentor senior engineers, lead design reviews, and drive the platform roadmap for next-generation GenAI workloads, embeddings, and retrieval-augmented systems.