Member of Technical Staff - GPU Performance Engineer

About the role

Reka, a globally distributed foundation model startup headquartered in the San Francisco Bay Area and working remotely first, is looking for an experienced GPU Performance Engineer with a strong background in Python and large-scale model training. You will design and implement improvements to the training infrastructure, take part in technical decisions that optimise model performance, and work on post-training processes such as reinforcement learning and fine-tuning.

Key responsibilities

• Design and implement improvements to the training infrastructure
• Contribute to technical decisions that optimise model performance
• Work on post-training processes, including reinforcement learning and fine-tuning
• Improve the efficiency and scalability of the model serving infrastructure

Required profile

• Strong engineering skills with fluency in Python and PyTorch (or other frameworks)
• Proven experience implementing and training large deep learning models
• Experience writing and debugging low-level GPU code (CUDA, C++)
• Experience scaling GPU jobs across large compute clusters (e.g. Slurm or Kubernetes)
• Ability to analyse and optimise GPU-accelerated workloads, including profiling, bottleneck identification and performance tuning

What we offer

• Collaboration with engineers, researchers and operators from organisations such as Google DeepMind and FAIR
• Access to cutting-edge infrastructure for training state-of-the-art models
• An open and inclusive culture
• Five weeks of paid leave and comprehensive healthcare benefits including vision and dental
• Visa support, including H1B and OPT transfers for US employees