Integrations

Works with the stack you already run

44 integrations across billing, Kubernetes, AI providers, metrics, notifications, identity and GitOps. The KubeOn agent reads your clusters, metrics and billing exports and never changes them; savings reach your clusters as pull requests in your own repository. Notifications, reports and exports send the data you choose, the UnitVane AI gateway routes the requests you send through it, and the opt-in GPU optimizer acts only within the policies you enable.

Request an integration

Integrations

Data readyUpdated 12 min ago

Last 7 days▾

Back to integrations

Connect cloud billing

CUR 2.0 data export, queried in place with Amazon Athena

Choose cloud
2Prepare export
3Grant access
4Connection details
5Test & save

Create the CUR 2.0 export

In the account that owns billing, turn on a daily CUR 2.0 export with resource IDs, and catalog it so Athena can query it.

CUR export bucket

acme-cur-exports

Billing console Copy

# 1. Billing preferences > Split cost allocation data for Amazon EKS: on
# 2. Data Exports > Create > Standard data export > CUR 2.0
#    - Include resource IDs: on
#    - Format: Parquet, delivered daily to s3://acme-cur-exports/
# 3. Catalog the export with an AWS Glue crawler

BackContinue

Connecting

Amazon Web Services

On save, KubeOn backfills 180 days of billing, then re-reads the last 3 days every morning.

Read-only access

KubeOn reads the billing export and nothing else in your account.

athena:StartQueryExecution
athena:GetQueryExecution
athena:GetQueryResults
glue:GetDatabase
glue:GetTable
s3:GetObject
s3:ListBucket

AWS CUR 2.0

Cloud billing

Data Export with resource IDs and split cost allocation data, queried in place with Amazon Athena.

Azure Cost Management

Cloud billing

Daily amortized or FOCUS exports, read with a managed identity.

Google Cloud Billing

Cloud billing

Detailed usage cost export in BigQuery, net of committed use discounts.

Oracle Cloud

Cloud billing

Cost and usage reports from Object Storage.

FOCUS 1.x

Cloud billing

Any FinOps Open Cost and Usage Specification file in object storage.

Data center rate cards

Cloud billing

Unit rates or amortized hardware cost for clusters you run yourself.

Amazon EKS

Kubernetes

EC2 node groups, Karpenter NodePools and Fargate.

Azure Kubernetes Service

Kubernetes

Node pools and virtual nodes, with Azure Workload Identity.

Google Kubernetes Engine

Kubernetes

Standard and Autopilot clusters, with GKE Workload Identity.

Red Hat OpenShift

Kubernetes

OpenShift 4 on any infrastructure, including ROSA and ARO.

Rancher RKE2

Kubernetes

Rancher-managed clusters in the cloud or the data center.

VMware Tanzu

Kubernetes

Tanzu Kubernetes Grid clusters on vSphere.

k3s and kubeadm

Kubernetes

Upstream Kubernetes on bare metal or virtual machines.

Karpenter

Kubernetes

Node sizing proposals written as NodePool requirement changes.

metrics-server

Metrics

Current pod CPU and memory usage, read at every snapshot.

Prometheus

Metrics

Hourly pod usage history and kubelet volume stats.

Thanos and Mimir

Metrics

Any Prometheus-compatible query API.

Managed Prometheus

Metrics

Amazon, Azure and Google managed Prometheus endpoints.

Slack

Notifications

Budget, alert, anomaly and digest messages to any channel.

Microsoft Teams

Notifications

Adaptive Cards through incoming webhooks or Workflows.

Email

Notifications

SMTP with STARTTLS, up to 50 recipients per channel.

Webhooks

Notifications

JSON payloads for PagerDuty, Opsgenie, ServiceNow or your own tools.

Okta

Identity

SAML 2.0 or OIDC sign-in with group-to-role mapping.

Microsoft Entra ID

Identity

SAML 2.0 or OIDC sign-in for your workforce tenant.

Google Workspace

Identity

OIDC sign-in with your Google domain.

Any SAML or OIDC provider

Identity

Standard SAML 2.0 and OpenID Connect.

Helm

GitOps and IaC

One chart for the agent, the same values in every cluster.

Flux

GitOps and IaC

HelmRepository and HelmRelease with per-cluster substitution.

Argo CD

GitOps and IaC

An Application per cluster, or an ApplicationSet for the fleet.

Terraform

GitOps and IaC

helm_release for agents, in the same apply as your clusters.

AWS CloudFormation

GitOps and IaC

The billing role for your payer account: it reads the export and writes only Athena query results.

Azure OpenAI

AI providers

Cost, tokens and per-minute PTU telemetry; a gateway target; the model behind Ask KubeOn on Azure.

Azure AI Foundry

AI providers

Foundry model deployments and their cost, alongside Azure OpenAI.

Amazon Bedrock

AI providers

On-demand and provisioned throughput cost and tokens; a gateway target.

Amazon SageMaker

AI providers

Endpoint cost by model and team from the CUR.

Google Vertex AI

AI providers

Gemini and partner model cost, tokens and provisioned throughput; a gateway target.

OpenAI API

AI providers

Organization cost and usage by project and model; a gateway target.

Anthropic API

AI providers

Workspace cost and usage by model; a gateway target.

Mistral

AI providers

La Plateforme usage and cost; a gateway target.

Self-hosted models

AI providers

vLLM and other servers on your GPUs, priced by UnitVane and routed by the gateway.

CSV and PNG

Data export

Any table as CSV across every page, any chart as PNG.

REST API

Data export

Allocation, assets and cloud costs with the UI's filters.

Scheduled reports

Data export

Daily, weekly or monthly delivery by email or Slack.

OpenTelemetry

Data export

Gateway request traces and metrics to your observability stack.

Ask for the integration you need.

Tell us what you run. Most integrations build on standard exports, APIs and webhooks.

Request an integration Read the docs