Roblox is hiring a
Machine Learning Infrastructure Engineer

As a Machine Learning Infrastructure Engineer you will build the next generation of ML Ecosystem Tooling. You will have an impact on the Roblox platform, and the industry, to the next level of managed E2E ML Pipeline Development and Automation—providing our developers and creators alike the ability to go from an ML idea to production in weeks or less. We are looking for accomplished ML Infrastructure and Systems Engineers to help build the next generation of ML Ecosystem Tooling.

You Are:

• Passionate about pushing the technological envelope and venturing into the unknown.
• Have a toolchest of system design experience upon which to draw to build scalable, reliable platforms for all of Roblox.
• 3+ years of professional experience working with scalable, distributed systems.
• Proficient in DevOps tooling such as Docker, K8S, CI/CD systems
• Have experience running and managing 1000+ node Kubernetes clusters
• Well versed with the Model Development Lifecycle from initial ad hoc analysis in notebooks to monitored services in production and back again.
• Experience with best practices in developing Platform / Infrastructure APIs
• Someone who has built a model in a modern ML framework such as Tensoflow, PyTorch, ML Lib
• Understand best practices around Data and Model management
• Have deployed and maintained an ML model in production
• Understand the use cases for multiple data processing and storage technologies
• Bachelor's degree in Computer Science, Computer Engineering, Data Science or a similar technical field.

You Will:

• Bootstrap and maintain infrastructure for ML Platform components--Serving Layer, Metadata Store, Model Registry, Feature Store, and Pipeline Orchestrator.
• Work to identify opportunities for further automation and optimization.
• Provide automation so our developers and creators alike the ability to go from an ML idea to production in weeks or less.
• Work on infrastructure projects such as GPU fleet management, workload optimization, and distributed training.
• Partner across organizations to build tooling, interfaces, and visualizations that make the ML@Roblox a delight to use.
• Have an impact to the state of the Roblox platform, and the industry, to the next level of managed E2E ML Pipeline Development and Automation
• Have an impact as a part of the team that is building a platform to handle the thousands of model experiments per day needed to support everything from ranking and recommendations, through content moderation and fraud prevention, to studio creative tooling.