Senior Principal AI Engineer
Core
Design and operate distributed training systems for large neural networks (autoregressive, diffusion, State Space Models) across GPU clusters to enable larger models and faster iteration cycles.
Role type
Senior Principal AI Engineer (Distributed Training Infrastructure)
Builds
GPU cluster orchestration, distributed communication layers, and large-model training pipelines for automotive voice AI.
Domain
Artificial Intelligence / Distributed Systems / High-Performance Computing
Deliverable
production ML models
Required skills
Distributed systems engineering, GPU cluster management, PyTorch distributed training, parallelism strategies (data/tensor/pipeline), low-level GPU communication and networking, memory optimization (activation checkpointing, ZeRO), fault tolerance at scale
Preferred skills
Experience with large language models or foundation models, HPC background
Technologies
Slurm, Kubernetes, Ray, RunAI, NCCL, RDMA, InfiniBand, NVLink, PyTorch Distributed, Megatron-LM, DeepSpeed
Responsibilities
Build and optimize GPU cluster orchestration; optimize and debug distributed communication; scale large-model training; apply advanced memory optimization techniques; ensure efficient scheduling, isolation, and fairness across training workloads
Seniority
Senior, hands-on IC
