Introducing Anyscale: The Future Is Distributed

By Robert Nishihara   

Announcing our $100M Series C and general availability of the Anyscale managed Ray offering.

Artificial intelligence has a problem.

From the board level down, companies know that AI is an urgent competitive necessity. But despite promises to redefine every industry, many companies struggle to get value out of AI. Underlying this struggle are a few core challenges: scale, production, and expertise.

It’s one thing to build a prototype on a laptop, it’s another to scale that prototype across hundreds of machines in the cloud. This transition is where so many teams get bogged down in building infrastructure and solving the tooling and scaling challenges adjacent to doing AI. Before you know it, doing machine learning means hiring an infrastructure team to build and maintain ad-hoc distributed systems. Meanwhile, the scale problem is only getting worse. OpenAI has estimated that compute demands for training state-of-the-art models have doubled every 3.4 months since 2012.

Moving from development to production is a similar story, and only 53% of AI projects successfully make this transition. It often involves a handoff to a separate team, it may involve rewriting the code using different frameworks, and it raises many more operational challenges around autoscaling, monitoring, and updating. Every step along the way can require custom tooling.

Because of these challenges, the success of machine learning initiatives today hinges on being able to hire experts. Unfortunately, there is simply not enough enterprise-grade AI engineering and distributed systems engineering to go around. And once the team is assembled, building your own infrastructure can easily take a year.

We started Ray and Anyscale because of our own experience struggling with the scaling challenges around machine learning. It’s striking to see how far the project has come, with industry leaders like Uber, OpenAI, Shopify, and ByteDance building their next generation ML platforms using Ray. Companies like McKinsey use Ray for AI-driven machine design to win competitions like the America’s Cup. Amazon processes petabytes of data per day and Ant Group deploys Ray on hundreds of thousands of cores.

Our goal is to enable every developer and every team to succeed with AI without needing to worry about building and managing infrastructure. We want to remove distributed systems expertise from the critical path of realizing the business potential of AI.

Over the past few months, we’ve been working closely with our beta customers to enable them to scale their AI applications and to bring AI to production, all with minimal code changes. Our customers can develop on their laptops and seamlessly scale in the cloud. They can prototype during development and deploy the same code in production, complete with production jobs, services, and monitoring. And they can do all of this without becoming infrastructure wizards.

We’ve seen organizations like mobile gaming giant, Wildlife Studios, and Swiss eCommerce leader, Ricardo, (who will be sharing their story on our virtual launch event next week) bring AI to production in record time and deploy hundreds of machine learning models, all without thinking about the underlying infrastructure.

Today, we’re thrilled to announce the general availability of Anyscale’s cloud platform. Our cloud platform will now be available to every developer and organization looking to scale AI. If you use Ray, Anyscale is the best way to run Ray.

Anyscale enables developers and teams to do a few core things:

  • Accelerate development and experimentation through instant access to cloud scale via an infinite laptop experience

  • Effortlessly deploy AI models and run AI workloads in production

  • Manage the end-to-end AI lifecycle on a single unified compute platform

  • Develop, scale, and deploy AI workloads without the hassle of managing infrastructure.

We are also delighted to announce our Series C funding round, a $100 million investment co-led by Andreessen Horowitz and Addition, along with participation from our existing investors. You can read more about the funding here.

If you’re looking to scale your AI applications or deploy machine learning in production, I encourage you to check out the Anyscale platform and join our virtual launch event next week to learn more.

And if you’re excited about enabling every developer and every organization to succeed with AI, come join us! We’re hiring for every role, and we are just at the very beginning.

Next steps

Anyscale's Platform in your Cloud

Get started today with Anyscale's self-service AI/ML platform:


  • Powerful, unified platform for all your AI jobs from training to inference and fine-tuning
  • Powered by Ray. Built by the Ray creators. Ray is the high-performance technology behind many of the most sophisticated AI projects in the world (OpenAI, Uber, Netflix, Spotify)
  • AI App building and experimentation without the Infra and Ops headaches
  • Multi-cloud and on-prem hybrid support