Services
Deploy your machine learning applications in production using Ray Serve, an open-source, distributed serving library for building online inference APIs.
Anyscale Services are a production-ready way to deploy Ray Serve applications, including key stability and performance features:
-
Fault tolerance: handle replica- and node-level failures without interruptions.
-
Zero downtime updates: safely update your applications using production-ready rollouts.
-
Autoscaling: use only the compute you need by dynamically adding and removing replicas in response to traffic.
-
Monitoring and observability: monitor service health and debug issues with an integrated UI, including log search, and metrics dashboards for key performance indicators.
Get started
- Sign in or sign up for an account.
- Select the Intro to Services example.
- Select Launch.
- This example runs in an Anyscale Workspace.
- Follow the notebook or view it in the docs.
- Terminate the workspace when you're done.
