Anyscale

Anyscale develops the Ray platform, an open-source distributed computing framework designed to make scalable AI and machine learning workloads accessible to ordinary developers. Ray runs on a single machine or across hundreds of nodes, allowing teams to scale from prototype to production without dedicated infrastructure engineering. The framework is used by OpenAI, Uber, Shopify and Amazon to power their machine learning platforms.

Beyond the open-source project, Anyscale offers an AI application development and deployment platform on which developers can build, debug and deploy AI applications using the same code throughout, from development to production and from one machine to hundreds of nodes. The design goal is a continuous path from local experimentation to large-scale deployment without rewrites.

The company's stated position is that scaling AI and machine learning should not require hiring specialised infrastructure engineers, and that developers should spend their time on problems rather than on managing infrastructure. Technical work spans distributed computing, machine learning infrastructure, distributed systems, developer tooling and MLOps, with an emphasis on reducing the friction between having an AI idea and deploying it.