Scalable Workflow Orchestration
Conductor is a platform originally created at Netflix to orchestrate workflows that span across microservices.
Open Source
Apache-2.0 license for commercial and non-commerical use. Freedom to deploy, modify and contribute back.
A fully abstracted backend enables you choose your own database persistence layer and queueing service.
Enterprise ready, Java Spring based platform that has been battle tested in production systems at Netflix and elsewhere.
Powerful flow control constructs including Decisions, Dynamic Fork-Joins and Subworkflows. Variables and templates are supported.
Client libraries in multiple languages allows workers to be implemented in Java, Node JS, Python and C#.
Distributed architecture for both orchestrator and workers scalable from a single workflow to millions of concurrent processes.
Developer Experience
  • Discover and visualize the process flows from the bundled UI
  • Integrated interface to create, refine and validate workflows
  • JSON based workflow definition DSL
  • Full featured API for custom automation
  • Understand, debug and iterate on task and workflow executions.
  • Fine grain operational control over workflows with the ability to pause, resume, restart, retry and terminate

Why Conductor?

Service Orchestration

Workflow definitions are decoupled from task implementations. This allows the creation of process flows in which each individual task can be implemented by an encapsulated microservice.

Designing a workflow orchestrator that is resilient and horizontally scalable is not a simple problem. Conductor was developed as a solution to that problem.

Service Choreography

Process flows are implicitly defined across multiple service implementations, often with tight peer-to-peer coupling between services. Multiple event buses and complex pub/sub models limit observability around process progress and capacity.