xelys jobs xelys jobs

Staff Software Engineer (ML Acceleration)

Stack AV

full-remoteleadpermanentbackenddataother Full remote 193 days ago via WTTJ

See how well this job matches your profile

Sign up to get an AI match score and generate a tailored application in seconds.

Get your match score

Tags

Machine LearningML OptimizationCUDAPyTorchTensorRTONNXGPU ProgrammingCustom GPU KernelsDeep LearningTriton

About the role

Role Overview

Join the ML Training Acceleration team at Stack AV as a Staff Software Engineer (ML Acceleration). You will help increase product development velocity by profiling, optimizing, and fine-tuning ML models, and by sharing best practices with Machine Learning Engineers.

Responsibilities

  • Profile, optimize, and fine-tune machine learning models to improve performance.
  • Collaborate with ML researchers to balance model accuracy and speed.
  • Implement ML and system optimizations to streamline end-to-end workflows.
  • Incorporate open-source tools into the ML acceleration pipeline.
  • Streamline model deployment (including conversion and runtime integration).
  • Evangelize best practices and frameworks across the company.

Requirements

  • Strong analytical and problem-solving skills.
  • CUDA programming experience.
  • Deep understanding of GPU architectures and performance optimization.
  • PyTorch optimization techniques.
  • Custom GPU kernel development.
  • TensorRT implementation.
  • ONNX model conversion and deployment.
  • Strong programming skills in C++ and Python.
  • Proven experience in GPU programming and optimization.
  • Excellent verbal and written communication; ability to explain technical concepts to non-technical stakeholders.
  • 5+ years of experience (including GPU programming/optimization).
  • Bachelor’s or Master’s degree in Computer Science, Engineering, or related field.

Nice to Have

  • Autonomous vehicles (AV) experience.
  • Familiarity with Triton for GPU kernels.

About Stack AV

Stack AV builds software products that accelerate machine learning workflows for faster product development. The company focuses on ML training and deployment performance, leveraging GPU-accelerated model optimization to improve iteration speed.

Scraped 5/13/2026