Scaling Multi-Modal Datasets to Petabytes with Ray at Apple

Learn how to scale multimodal data processing frameworks to handle petabyte-scale workloads from Apple engineers who share their real-world experience at Ray Summit 2025. Discover the rising demand for large-scale processing over multimodal datasets and understand the challenges of designing systems that balance user-friendliness with extreme scalability capabilities. Explore how Ray's distributed computing model provides flexibility while requiring careful attention to bottlenecks across data processing layers, Ray's core runtime, and underlying infrastructure to achieve petabyte-level performance. Examine key strategies for mitigating performance bottlenecks, ensuring infrastructure resilience, and tuning Ray's internal components to prevent cascading failures in massive, tightly coupled distributed systems. Gain practical insights from architecting and productionizing Apple's petabyte-scale multimodal pipeline as a self-serve service that balances usability with the ability to process enormous datasets reliably and efficiently. Walk away with actionable knowledge for building and operating large-scale multimodal data workflows in real-world production environments as LLM applications continue to grow in complexity.