Announcing the Ray Foundations Certification
Ray has become the distributed compute framework that unifies the AI ecosystem and its being used by industry leaders to build and run AI at scale. Examples include Attentive, which is building multimodal personalization models, and Shopify, which is leveraging vision LLMs to revamp e-commerce catalogs and the list goes on. As Ray adoption continues to exponentially grow, demand is rising for researchers, developers and engineers who can apply it effectively.
That’s why we’re excited to launch the Ray Foundations Certification. This credential is designed to validate your ability to work with Ray’s core architecture, primitives, and libraries.
Ready to prove your knowledge? Access certification exam here.
LinkWhy Earn This Certification?
Ray adoption is accelerating: AI-first organizations like Uber, Pinterest, Canva, Nubank and hundreds of others already rely on Ray as the foundation for their most demanding AI workloads. As Ray becomes the standard for distributed AI and scalable compute, expertise in Ray is rapidly turning into a requirement to be part of some of the leading AI teams around the world.
Figure 1. Ray open source project GitHub star growth. Generated with star-history.com
Career growth: A certification in Ray signals that you have the necessary foundational skills to build scalable, production-ready AI systems. Distributed computing skills are becoming essential, and Ray is at the center of this shift.
Figure 2. Sample ML Engineer LinkedIn Job Postings qualifications and skillsLinkWho Is It For?
This certification is designed for builders early in their Ray journey who want to demonstrate essential distributed computing skills:
ML Engineers and Data Scientists building and scaling multimodal data prep and distributed training pipelines.
Data Engineers working on preprocessing, batch inference, and large-scale pipelines for unstructured data.
Platform/Infra Engineers supporting the broad spectrum of AI - from ML to Agentic AI - in production.
AI Researchers experimenting with new models or techniques who need to interactively manipulate massive datasets and run large-scale experiments.
No prerequisites are required, just curiosity and a drive to upskill.
LinkRecommended Trainings
To prepare for the exam, we recommend taking these self-paced, free courses:
Batch Inference and Data Processing with Ray Data. Link
Distributed Training with Ray Train. Link
Hyperparameter Tuning with Ray Tune
Online Model Serving with Ray Serve. Link
Full details and other recommended reading can be found in the exam guide.
LinkGet Certified
The exam includes 60 multiple-choice questions, takes 120 minutes, and is delivered online. A passing score is 70 percent. It covers Ray clusters and architecture, Ray Core, Ray Data, Ray Train, Ray Tune, and Ray Serve.
