Jobs in Palo Alto, CA

About the role

We are hiring Research Engineers to sit at the boundary of research and systems engineering. Our multimodal data spans trillions of tokens of video alongside rich neural and behavioral recordings, making this one of the most demanding dataloading challenges in frontier AI. You will own the systems that turn this large, heterogeneous data into training-ready multimodal streams for foundation model training and evaluation at scale. This means designing and building a state-of-the-art end-to-end dataloading stack: data formatting, preprocessing, filtering, sharding, caching, and streaming. You will build runtime interfaces that deliver data to distributed training jobs across GPU clusters with high throughput, reliability, and full observability. You'll have substantial autonomy to shape foundational technical decisions on a small, high-impact team. We are hiring Research Engineers to sit at the boundary of research and systems engineering. You will own the systems that turn this large, heterogeneous data into training-ready multimodal streams for foundation model training and evaluation at scale. This means designing and building a state-of-the-art end-to-end dataloading stack: data formatting, preprocessing, filtering, sharding, caching, and streaming. You will build runtime interfaces that deliver data to distributed training jobs across GPU clusters with high throughput, reliability, and full observability.

What you'll bring

• Bachelor's degree or equivalent experience in Computer Science, Machine Learning, Computational Neuroscience, or a related field
• Strong software engineering skills in Python and familiarity with PyTorch
• Experience building high-throughput data pipelines or dataloading systems for large-scale distributed ML training
• Experience working with and building systems for complex, multimodal time-series data
• Experience with video processing at scale: decoding, transcoding, I/O optimization for large video corpora
• Hands-on experience profiling and benchmarking data systems on metrics such as throughput, IOPS, GPU utilization, and memory usage
• Familiarity with multi-modal transformer architectures
• Experience with distributed training environments and deep understanding of sharding models and data
• Experience with containerization, and scaling container orchestration (e.g. via Docker, Kubernetes)
• Background in scientific computing, computational neuroscience, life sciences, or ML-adjacent research environments
• If you don't see a role that fits, we encourage you to submit a general application and tell us how you'd like to contribute to our mission.
• Experience with ML workflow orchestrators (e.g. Prefect, Dagster, Airflow).