Senior Principal Software Engineer - Ray/KubeRay

About the job

Foundation Models and Large Language Models (LLMs)? The OpenShift AI team is looking for a Principal Software Engineer with Kubernetes and MLOps experience to join our rapidly growing engineering team. Our team creates the tools necessary for enterprise data science teams to leverage distributed compute infrastructure across the hybrid cloud when training today’s most complex ML models. This is a very exciting opportunity to build and impact creation of the next generation of AI.

 

In this role, you'll be contributing as an expert on ML model training and the tools necessary to support it at enterprise scale. Your contributions will enable bleeding edge platform capabilities around distributed computation and model training, kubernetes-native job scheduling, and kubernetes-native resource optimization. You’ll lead the architecture and implementation of this space in the open source Open Data Hub project by actively participating in strategic upstream communities like Ray/KubeRay, PyTorch, HuggingFace, and others. You will work as part of an evolving development team to rapidly design, secure, build, test and release distributed model training capabilities. The role is primarily an individual contributor who will be a key notable contributor to MLOps upstream communities and collaborate closely with the internal cross-functional development teams. 

What you will do

• Be an influencer and leader in MLOps related open source communities to help build an active MLOps open source ecosystem for Open Data Hub and OpenShift AI

• Architect and lead implementation of scalable open source solutions for Data Scientists to leverage distributed computing capabilities to train their Machine Learning models, running on OpenShift

• Act as a MLOps SME within Red Hat by supporting customer facing discussions, presenting at technical conferences, and evangelizing OpenShift AI within the internal community of practice

• Architect and design new features for open source communities such as KubeRay, Ray, PyTorch, and CodeFlare

• Provide technical vision and leadership on critical and high impact projects 

• Mentor, influence, and coach a distributed team of engineers

What you will bring

• An existing contributor in one or more MLOps open source projects such as Ray/KubeRay, KubeFlow, Pytorch, or Spark

• Recent hands on experience with distributed computation tools like Ray or Spark, either at the end-user or infrastructure provider level

• Experience training ML models using tools like Pytorch or Tensorflow

• Advanced level of experience with Kubernetes

• Advanced level knowledge and experience in development in Go and Python

• Technical leadership acumen in a global team environment

• Passion for writing and maintaining reliable code

• Excellent written and verbal communication skills; fluent English language skills

The following will be considered a plus: 

• Bachelor's degree in statistics, mathematics, computer science, operations research, or a related quantitative field, or equivalent expertise; Master’s or PhD is a big plus

• Experience in engineering, consulting or another field related to distributed model training or data processing in a customer environment or supporting a data science team

• Highly experienced in OpenShift

• Familiarity with popular python machine learning tools such as PyTorch, Tensorflow, and Hugging Face

 

 

#LI-LS2

#LI-REMOTE