AI Tooling Architect for Efficient Training & Inference

Jobtailor; Sunnyvale (CA); On-site; USD 140,000 - 190,000; Full time

Job summary

Jobtailor in Sunnyvale seeks an experienced AI/ML engineer to enhance training and inference efficiency in large-scale models. You will build workflows for correctness and performance and create tooling to predict model behavior from architecture and data shapes.

Proficiency with PyTorch, TensorFlow, TensorRT and NVIDIA Nsight tools is required, as is production-grade Python/C++ coding. You will influence architecture decisions within GM and help evolve the AI toolchain.

Qualifications

• 3+ years of AI/ML experience.
• Experience with PyTorch and TensorFlow.
• Proficiency writing production-grade Python/C++ code; SDLC.
• BS or higher in CS/CE/EE, or equivalent.

Responsibilities

• Identify opportunities to improve training and inference efficiency.
• Build workflows for correctness and performance on in-car compute and sensors.
• Develop tooling to predict model performance from architecture and data shape.
• Trace performance on large distributed training/distillation on GPUs.
• Evolve toolchain to leverage advancements in AI.
• Influence model architecture decisions within GM.

Skills

Python

C++

AI/ML

Production code

SDLC

Education

Bachelor's degree in CS/CE/EE

Tools

PyTorch

TensorFlow

TensorRT

Nsight-systems

Nsight-compute

Job description

Jobtailor in Sunnyvale seeks an experienced AI/ML engineer to enhance training and inference efficiency in large-scale models. You will build workflows for correctness and performance and create tooling to predict model behavior from architecture and data shapes.

Proficiency with PyTorch, TensorFlow, TensorRT and NVIDIA Nsight tools is required, as is production-grade Python/C++ coding. You will influence architecture decisions within GM and help evolve the AI toolchain.