Adoption & Ecosystem

Who uses it

The CNCF graduation announcement names these production adopters ([4]). The core repository has no ADOPTERS.md; the listed-user roster is maintained on the keda.sh site ([1], README.md:66-68).

Organisation	Use case	Source
FedEx	Production KEDA user named at graduation	[4]
Grafana Labs	Production KEDA user named at graduation	[4]
KPMG	Production KEDA user named at graduation	[4]
Reddit	Production KEDA user named at graduation	[4]
Xbox	Production KEDA user named at graduation	[4]
Zapier	Production KEDA user named at graduation	[4]

The same announcement states that 45+ organisations run KEDA in production ([4]).

Adoption signals

Observed on 2026-06-22 via gh api repos/kedacore/keda ([1]):

• GitHub stars: 10,310
• Forks: 1,441
• Contributors: roughly 455

At graduation the project reported 60+ scalers and 9 authentication providers ([4]); the pinned source has roughly 78 cases in the scaler build switch ([2], pkg/scaling/scalers_builder.go:123).

Ecosystem

• HPA: native integration through the Kubernetes External Metrics API; KEDA creates and manages the HPA object.
• Event sources: 70+ built-in scalers including Prometheus, Kafka, RabbitMQ, NATS, AWS SQS/Kinesis/CloudWatch, Azure Service Bus/Event Hub/Queue, and GCP Pub/Sub ([2], pkg/scaling/scalers_builder.go:123).
• Authentication: TriggerAuthentication with Vault, Azure Key Vault, and Pod Identity ([2], apis/keda/v1alpha1/triggerauthentication_types.go:75).
• HTTP add-on: kedacore/http-add-on extends KEDA to scale HTTP workloads on request volume ([8]).
• Downstream: OpenShift ships KEDA as its Custom Metrics Autoscaler in openshift/kedacore-keda ([9]).

Alternatives

Alternative	Differs by
Native HPA + a metrics adapter (e.g. Prometheus Adapter)	KEDA bundles 70+ scalers and adds scale-to-zero; the bare HPA cannot go below one replica
Knative Serving	Strong at HTTP/request-driven scale-to-zero; KEDA covers a broader set of generic event sources
Cluster Autoscaler / Karpenter	These scale nodes; KEDA scales pods, so they operate at different layers and are commonly used together
Cloud-specific managed autoscalers	KEDA is vendor-neutral across clouds ([4])