The Role
As a member of technical staff, you will lead the core engineering effort to build compute efficient, physically accurate world models. Your work will enable our AIs to "dream" and predict how the world evolves, and help engineers evaluate and debug physical AI systems in real-like environments.
You are responsible for the complete data lifecycle of the World Model, from multimodal sourcing and synthetic generation to high-throughput GPU delivery. You have a deep understanding of the data infrastructure stack, and you are capable of delivering feedback and technical contributions at every layer of the data-to-compute pipeline.
Job Responsibilities
- Area: Design and scale the systems for large-scale data ingestion, qualification, and high-performance delivery to the compute clusters.
- Ownership: The "Data-to-Compute" pipeline—ensuring GPUs are never starved for data and the I/O subsystem can saturate massive-scale training runs.
- Focus: Building high-speed data loaders and distributed retrieval systems that eliminate the bottleneck between the data lake and the GPU.
- Operations: Optimize storage architectures, caching strategies, and streaming protocols to maximize compute utilization and minimize research turnaround time.
Minimum Qualifications
- Extensive experience in software engineering for data-intensive deep learning R&D teams.
- Expert-level experience architecting large-scale data systems, high-performance dataloaders, and distributed storage for AI training.
- Expert with Python, C++ / Rust.
Preferred Qualifications
- Proof of outstanding contribution to open-source libraries for large-scale data orchestration, simulation, or high-performance I/O.
- Peer-reviewed research contribution in deep learning (e.g., data-centric AI, world models, or simulation) from a AAA conference.