ML Performance Engineer (AI)

TL;DR

ML Performance Engineer (AI): Optimizing large-scale model training by enhancing the internal stack and compute infrastructure with an accent on GPU-level optimization and performance tuning. Focus on identifying bottlenecks, building tools for high-performance workflows, and mentoring team members.

Company

hirify.global is focused on building deep technical expertise in ML training systems.

What you will do

• Optimize model training pipeline for speed and reliability.
• Apply GPU-level optimization techniques using JAX, Triton, and CUDA.
• Identify and resolve performance bottlenecks across the ML pipeline.
• Build tools and extend internal infrastructure for scalable training workflows.
• Mentor engineers and researchers in performance best practices.
• Contribute to a culture of engineering excellence and rapid experimentation.

Requirements

• Experience optimizing neural network training in production or research settings.
• Practical experience with ML frameworks like PyTorch or JAX.
• Hands-on experience with deep learning architectures such as LSTM and Transformers.
• Experience with CUDA, Triton, or low-level GPU technologies.
• Proficiency in profiling and debugging training pipelines.
• Understanding of distributed training concepts.
• Strong proficiency in Python for building infrastructure-level tooling.