Frequently Asked Questions
General
What is RubixKube?
RubixKube is an AI-native Site Reliability Intelligence (SRI) platform. It goes beyond traditional monitoring by using specialized AI agents to detect anomalies, diagnose root causes, and recommend or execute remediation actions across your infrastructure.What environments do you support?
RubixKube supports Kubernetes (EKS, GKE, AKS, bare metal), AWS, GCP, Azure, and Linux VMs. You can manage multiple clusters and environments from a single tenant workspace.What browsers are supported?
The RubixKube Console supports the latest versions of Chrome, Edge, Safari, and Firefox.Setup & Installation
How do I connect my first cluster?
Navigate to Clusters → Connect Cluster in the Console. You will be provided with akubectl apply command to install the Observer agent. Once deployed, it connects automatically to the RubixKube control plane via NATS JetStream.
How long until data shows up?
Data typically appears within 1–2 minutes after the Observer agent is successfully installed, depending on the size of your cluster and the volume of telemetry.Can I manage multiple clusters?
Yes. Multi-cluster support is available on Pro and Enterprise plans. You can switch between clusters using the Cluster Selector in the Console header, and Insights/RCA reports will be scoped to the selected cluster.AI Agents & RCA
How does the AI detect incidents?
The Observer agent continuously processes real-time telemetry (events, logs, metrics) within your cluster to identify anomalies. It then sends summarized and anonymized data (Insights and infrastructure snapshots) to the Memory Engine. The Memory Engine uses fingerprinting and state machines to build context, deduplicate issues, and trigger the RCA pipeline without ever receiving a raw dump of your logs or metrics.What agents are available?
RubixKube features multiple specialized agents:- SRE Agent: The primary agent for troubleshooting, analyzing Kubernetes graphs, logs, and metrics.
- Prism (Executive Agent): Provides high-level, BLUF-style (Bottom Line Up Front) summaries for leadership.
- Expert SRE: Conducts deep, specialized reliability analysis.
Can I trust the Root Cause Analysis (RCA)?
Yes. Every RCA report includes a confidence score and direct links to the underlying evidence (logs, metrics, events) so your engineers can validate the AI’s conclusions before taking action.Integrations & Actions
What integrations are supported?
RubixKube supports native integrations with Slack, PagerDuty, GitHub, GitLab, Linear, Notion, and major observability tools (Datadog, New Relic, Prometheus). We also support Custom MCP (Model Context Protocol) servers for extending agent capabilities.How do Actions work?
When an RCA identifies a fix, it can propose an Action. Actions follow a strict state machine (todo → assigned → in_progress → review → completed). Depending on your Guardrails configuration, actions can be executed automatically or require human approval.
Billing & Security
Do you store application data?
No. The Observer agent only collects infrastructure metadata, events, and aggregated metrics. It does not inspect or store your application payloads or sensitive customer data.What authentication methods are supported?
We use Auth0 for secure authentication, supporting SSO (Google, GitHub) as well as standard email/password login. Enterprise plans include advanced SSO and SAML configurations.How is my data isolated?
RubixKube employs strict multi-tenancy. Every API request, database record, and NATS message is securely scoped by a uniqueTenant ID.
Getting Help
Where can I get more help?
- Check the Troubleshooting Guide for common issues.
- Use the Rubix Chat Agent in the Console for interactive assistance.
- Contact support at connect@rubixkube.ai with your Tenant ID, repro steps, and screenshots.