What is
OpenChoreo
?
OpenChoreo is an open source platform orchestrator for building internal developer platforms on Kubernetes. Its control plane turns developer intent declared through APIs into running applications, infrastructure resources and AI agents, and orchestrates separate data, workflow and observability planes to do so.
At a glance
Attribute | Value |
|---|---|
Primary use case | Orchestrate an internal developer platform on Kubernetes for applications, infrastructure resources and AI agents |
Deployment type | Self-hosted on Kubernetes; control plane plus data, workflow and observability planes on one or many clusters |
Pricing model | Open source; commercial enterprise support from WSO2 (pricing not published) |
API compatibility | Kubernetes-native declarative APIs (CRDs), OpenAPI-based REST API, occ CLI, MCP servers |
Supported integrations | Kubernetes, Backstage, Argo Workflows, GitHub Actions, Jenkins, Argo CD, Flux, OpenSearch, Prometheus, OpenTelemetry, Cilium, OAuth2 and OIDC identity providers |
Source type | Open |
Cloud provider | Cloud Agnostic |
AEP branch | Tooling |
AEP category | Integration & Delivery |
AEP subcategory | IaC & Platform Orchestration |
Profile
OpenChoreo is a platform orchestrator: its control plane is the central component that reconciles the desired state of platform and developer resources with the other planes. Developers and platform engineers declare intent through Kubernetes-native APIs (services, web apps, dependencies, environments), and the control plane decides what is deployed, where and how. Data planes run the workloads, workflow planes run CI, GitOps and other automation workflows, and observability planes collect logs, metrics and traces. Each plane can be installed, scaled and upgraded on its own, on one cluster or many.
The same control plane is used through a Backstage-based portal, the occ CLI, an OpenAPI-based API and MCP servers, all under one authorization model built on Apache Casbin. Tools behind the planes are modules that can be swapped, for example Argo Workflows for CI, OpenSearch for logs and Prometheus for metrics.
Focus
Platform teams that want to build an internal developer platform on Kubernetes without assembling and integrating every tool themselves, and that need it to serve both developers and AI agents under the same guardrails. Enterprises with sovereignty requirements can self-host it on any Kubernetes infrastructure.
Background
WSO2 developed OpenChoreo as the successor to its SaaS WSO2 Developer Platform, formerly WSO2 Choreo, and released it as open source. It is a CNCF Sandbox project. WSO2's commercial model is enterprise support and architecture guidance.
Features
Control plane as orchestrator
The control plane takes intent declared through the API and reconciles it across data, workflow and observability planes, so developers work with services, environments and dependencies rather than Kubernetes resources.
Independent planes
Data, workflow and observability planes are separate deployable units with their own lifecycle. Teams can scale or upgrade one plane, or run several per region or environment, without a platform-wide upgrade.
Swappable modules
Defaults such as Argo Workflows for CI, OpenSearch for logs and Prometheus for metrics can be replaced with existing tools, and external CI such as GitHub Actions or Jenkins can be plugged in, without changing the developer experience.
Cells for runtime isolation
Each project and environment runs in a Cell with a defined gateway topology for ingress and egress, giving zero-trust isolation between components. An optional module uses Cilium and eBPF to enforce network policies.
Many interfaces, one access model
A Backstage-based portal, the occ CLI, the API and MCP servers all act on the same control plane, with role- and attribute-based access control that also applies to AI agents.
Built-in platform agents
SRE and FinOps agents provide root cause analysis, remediation and cost optimization, using the observability plane's MCP server.
Incremental adoption
Existing clusters can be connected read-only first, with workloads mapped into the OpenChoreo model, before OpenChoreo takes over deployments.
Pros and cons
Pros
Gives platform teams a working orchestrator with control, data, workflow and observability planes instead of integrating each tool from scratch.
Swappable modules let teams keep tools they already run behind a consistent developer experience.
Humans and AI agents use the same API, CLI, portal and MCP servers under one access model.
Planes can be scaled and upgraded independently and run across one or many clusters.
Open source and self-hosted on any Kubernetes infrastructure, which suits sovereignty requirements.
Cons
Runs only on Kubernetes and requires the team to operate it, including Kubernetes administration and GitOps skills.
A young project: CNCF Sandbox status, with general availability in 2026, so production references are still few.
Its breadth overlaps with tools teams may already use for portals, CI and observability, so adoption means deciding which roles OpenChoreo takes over.
FAQs
What is OpenChoreo?
OpenChoreo is an open source platform orchestrator for building internal developer platforms on Kubernetes. Its control plane takes developer intent declared through its APIs and orchestrates separate data, workflow and observability planes to deliver it.
Is OpenChoreo a developer portal?
No. The portal is one interface to the orchestrator. OpenChoreo ships a Backstage-based portal, but the same control plane is also used through the occ CLI, the API and MCP servers, all with the same access controls.
Who develops OpenChoreo?
WSO2 created OpenChoreo as the successor to its SaaS WSO2 Developer Platform, formerly WSO2 Choreo. It is a CNCF Sandbox project, and WSO2 sells enterprise support and architecture guidance for it.
Can OpenChoreo work with existing tools and clusters?
Yes. Its default modules, such as Argo Workflows for CI and OpenSearch for logs, can be replaced with tools teams already run. Existing clusters can first be connected read-only, with workloads mapped into the OpenChoreo model, before OpenChoreo manages deployments.
Does OpenChoreo support AI agents?
Yes. The control plane and the observability plane each expose an MCP server, and agents work under the same role- and attribute-based permissions as the developer they act for. Built-in SRE and FinOps agents handle root cause analysis, remediation and cost optimization.
How much does OpenChoreo cost?
OpenChoreo is open source and self-hosted. WSO2 offers commercial enterprise support; pricing for it is not published.

