Senior AI Infra Engineer — High-Throughput Multimodal ML

Kodiak; Mountain View (CA); On-site; USD 190,000 - 260,000; Full time

Benefits offered by this job

Competitive compensation

Equity

Medical/Dental/Vision

Flexible PTO

Mountain View office

Lunch provided

EV charging

Wellbeing benefits

401(k)

Job summary

Kodiak Robotics, Inc. is seeking engineers to accelerate model training for autonomous driving data pipelines in Mountain View, CA. The role focuses on building high‑throughput data loading, scalable distributed training, and GPU‑level optimizations to saturate modern accelerators.

You will work with PyTorch, DeepSpeed, and related tools to scale multimodal sensor data across multi‑node GPU clusters, ensuring efficient, reliable training performance for large autonomous driving models.

Qualifications

• BS, MS, or PhD in Computer Science or a related field, and at least 2–3 years of industry experience in ML systems or infrastructure.
• Experience with distributed training frameworks and techniques (PyTorch DDP/FSDP, DeepSpeed, Megatron, NCCL) and a strong grasp of parallelism trade-offs.
• Experience building high‑performance data pipelines for large‑scale training, including streaming dataset formats (WebDataset, MosaicML Streaming/MDS, or similar), sharding, and storage/network‑aware loading.

Responsibilities

• Design high-throughput data loading and streaming systems for multimodal sensor data (camera, LiDAR, radar), including dataset formats, sharding strategies, and prefetching pipelines that keep GPUs saturated
• Build and optimize distributed training infrastructure across multi‑node GPU clusters, applying data, tensor, pipeline, and fully sharded (FSDP/ZeRO) parallelism to models that don't fit on a single device
• Maximize utilization of modern accelerators such as NVIDIA B200s through mixed‑precision training (BF16/FP8), fused kernels, memory optimization, and communication/computation overlap
• Profile end‑to‑end training pipelines to find and eliminate bottlenecks across storage, network, CPU preprocessing, and GPU compute
• Develop scalable dataset construction pipelines that convert petabytes of raw driving logs into training‑ready, streamable formats
• Partner with ML teams to scale new architectures from prototype to full‑cluster training runs efficiently and reliably

Skills

Distributed training

Data pipelines

GPU optimization

Python/PyTorch

C++/CUDA/Triton

Education

BS/MS/PhD in Computer Science or related field

Tools

WebDataset

MosaicML Streaming

NCCL

NVLink

InfiniBand

Job description

Kodiak Robotics, Inc. is seeking engineers to accelerate model training for autonomous driving data pipelines in Mountain View, CA. The role focuses on building high‑throughput data loading, scalable distributed training, and GPU‑level optimizations to saturate modern accelerators.

You will work with PyTorch, DeepSpeed, and related tools to scale multimodal sensor data across multi‑node GPU clusters, ensuring efficient, reliable training performance for large autonomous driving models.