Job Summary
The role involves designing and constructing robust, scalable multimodal training input pipelines starting from certified game capture, through data processing, and ultimately delivering the data to various accelerators. The position emphasizes establishing standards for pipelines and performance benchmarks, diagnosing issues across layers, and guiding architecture evaluations while also providing mentorship to engineers. Collaboration with research and infrastructure teams is essential.
Responsibilities
- Architect and develop scalable multimodal training input pipelines.
- Establish and enforce pipeline standards and performance goals.
- Identify and resolve cross-layer bottlenecks and failures.
- Conduct architecture reviews and mentor engineering staff.
- Collaborate with research and infrastructure teams to enhance pipeline performance.
Qualifications
- A minimum of 20 years of development experience.
- Proven experience in managing production distributed-training input systems.
- Demonstrable improvements in performance, reliability, recovery, or cost efficiency.
- Expert knowledge of large-scale training frameworks and accelerator parallelism.
- Strong skills in multimodal data ingestion, dataset management, deterministic and fault-tolerant loading, and proficiency in profiling distributed systems.
Electronic Arts provides equal employment opportunities to all employees and applicants for employment and prohibits discrimination and harassment of any type without regard to race, color, religion, age, sex, national origin, disability status, genetics, protected veteran status, sexual orientation, gender identity or expression, or any other characteristic protected by federal, state or local laws.
