rhoai-architecture-overview

name	rhoai-architecture-overview
metadata	{"author":"rhoai3-demo","version":"1.0.0","platform-family":"rhoai","platform-baseline":"repo","ocp-baseline":"repo","skill-group":"RHOAI Platform"}
description	Use when describing, reviewing, or implementing the high-level Red Hat OpenShift AI Self-Managed architecture from the official architecture chapter: service-layer capabilities, the Red Hat OpenShift AI Operator management layer, default projects/namespaces, custom project boundaries, and the rule not to place ISV applications in OpenShift AI namespaces. Also use when creating README architecture narratives or GitOps namespace layouts that need to align with official RHOAI architecture. Do NOT use for specific component CR fields, operator install commands, or live troubleshooting; use the relevant component skill, official docs, and env-* skills.

RHOAI Architecture Overview

Use this skill to ground project architecture narratives and namespace topology in the official Red Hat OpenShift AI Self-Managed architecture chapter for the active baseline in docs/PLATFORM_BASELINE.md.

Source Grounding

Read references/source-capture.md before using product configuration details. Official Red Hat documentation is product authority. This skill is high-level: it identifies layers, included services, default projects, and namespace boundaries, but it does not define component CR fields.

Architecture Model

Use the official two-layer framing:

• Service layer: OpenShift AI dashboard, model serving, AI pipelines, self-managed Jupyter, distributed workloads, and RAG through the integrated Llama Stack Operator.
• Management layer: Red Hat OpenShift AI Operator as the meta-operator that deploys and maintains OpenShift AI components and sub-operators.

When writing README or GitOps architecture content, keep custom demo services separate from product services. Label custom glue explicitly.

Namespace Guidance

For predefined projects, align with the official namespace roles:

• redhat-ods-operator: Red Hat OpenShift AI Operator.
• redhat-ods-applications: dashboard and other required OpenShift AI components.
• rhods-notebooks: default project for basic workbenches.

Additional application projects are expected for workloads that use machine learning models. Do not install independent software vendor applications in namespaces associated with OpenShift AI.

Workflow

1. Confirm the active baseline in docs/PLATFORM_BASELINE.md.
2. Read references/official-doc-extraction.md.
3. For architecture narratives, map every claimed capability to the official service-layer or management-layer component.
4. For GitOps namespace layout, keep product namespaces and demo application namespaces separate.
5. For implementation-specific fields or install commands, switch to the relevant component skill or official docs chapter.
6. Validate with references/validation-checklist.md.

References

• references/source-capture.md
• references/official-doc-extraction.md
• references/validation-checklist.md
• examples/namespace-topology.md