Staff Software Engineer, Ray Data

The role

You will design, build, and improve core systems for Ray Data. This includes optimizing distributed execution and building data loading solutions for AI workloads.

What you will do

• Design, build, and improve core systems for Ray Data.
• Design and optimize distributed execution across data pipeline stages.
• Build data loading and processing solutions for production training and inference workloads.
• Solve problems in distributed execution, scheduling, resource management, data partitioning, fault tolerance, and performance optimization.
• Make system-level architectural decisions regarding resource allocation, execution models, and consistency.
• Work with customers and AI-native companies to solve scaling challenges.

Required qualifications

• 6+ years of experience building production-grade software, infrastructure, or developer-facing systems.
• Strong Python engineering experience.
• 6+ years of experience personally owning core architectural decisions within a distributed data or compute engine.
• Deep experience with distributed systems internals such as scheduling, fault tolerance, data partitioning, distributed execution, performance optimization, or database and query engine internals.
• Track record of reasoning through system-level tradeoffs and defending architectural decisions.

Nice to have

• Passion for solving the unsolved problems in large-scale AI infrastructure and building systems that enable the next generation of AI applications.

Benefits

• Competitive salary and equity
• Health/dental/vision coverage
• Flexible time off
• Paid parental leave
• Mental health support

About the company

Anyscale is commercializing Ray, an open-source project for scalable machine learning. The company is backed by Andreessen Horowitz, NEA, and Addition.