Red Hat OpenShift AI

Learn how to use the open source scanner garak to test AI models for

Learn how Red Hat built IdeaBot on OpenShift using isolated pods, auth proxies, and EARS-based BDD to secure and automate internal AI research.

Learn how to add NeMo Guardrails to LangGraph agents on Red Hat OpenShift AI using a proxy pattern to inspect requests without modifying agent code.

Compare Jev decision models with Red Hat guardrails to evaluate AI safety performance, cost, and latency for enterprise implementation.

Learn how to unify Kubernetes and OpenShift authentication with Kube AuthKit, a Python library that simplifies token management across environments.

Learn how to simplify distributed training with Kubeflow Trainer 2 on OpenShift AI by replacing framework-specific CRDs with a unified TrainJob resource.

Discover how to build cost-ranked fraud detection models in Python, combine predictive ML with LLM triage notes, and scale using Red Hat OpenShift AI.

Add NeMo Guardrails to a LangGraph agent on Red Hat OpenShift AI without rewriting the graph. Jump to 5:28 to see the same in-cluster vLLM unguarded vs guarded: the unguarded agent answers a check-fraud prompt; NeMo Guardrails blocks it on content safety.
NeMo Guardrails is included in Red Hat OpenShift AI (upstream: NVIDIA NeMo Guardrails) and is deployed with a NeMo Guardrails custom resource managed by the TrustyAI Operator. This walkthrough covers the passthrough proxy, layered rails (regex → content safety → topic boundary → output safety), and opt-in tracing so you can see which rail fired.

Chapters
0:00 Introduction
0:17 What this agent does
1:02 Proxy architecture and rail order
2:25 Customizing rails and changing domain
4:23 Deploy locally and on OpenShift AI
4:58 Tracing overview
5:28 Unguarded vs guarded
8:04 Which rail fired
9:35 Wrap-up

Resources

Guardrailed Agent example:
https://github.com/red-hat-data-services/agentic-starter-kits/tree/main/agents/langgraph/examples/guardrailed_agent

Enable AI safety with NeMo Guardrails:
https://docs.redhat.com/en/documentation/red_hat_openshift_ai_self-managed/3.5/html/enabling_ai_safety_with_guardrails/enabling-ai-safety-with-nemo-guardrails_nemo-guardrails

Track AI inference costs by department using Red Hat OpenShift AI’s built-in MaaS gateway API keys, Perses dashboards, and MLflow tracing, with no manual instrumentation required.

Learn how to restrict, verify, and control agent actions with Red Hat OpenShift AI.

Automate enterprise-wide RAG pipelines with Red Hat OpenShift AI's AutoRAG, speeding up AI optimization for your business.

This article demonstrates how to deploy a NeMo Guardrails configuration on a Red Hat OpenShift AI cluster. It covers prerequisites, the deployment process, and testing the deployed server.

Learn how to fine-tune your text-to-speech model for Turkish with Red Hat OpenShift AI and Kubeflow Trainer, reducing speech errors by over 90%.

Improve large language model inference speed with Speculators 0.6.0's FastMTP-style fine-tuning.

Evaluate LLM guardrails with EvalHub: Improve AI safety with open-source platform

Evaluate AI agents on Red Hat OpenShift AI with IBM CLEAR and EvalHub. Learn how to automate analysis for recurring failure patterns in AI agent execution.

Learn how the vLLM LoRA dynamic loading flaw enables data theft and how to enforce zero trust defenses using Red Hat OpenShift AI and cluster management.

Learn how to automate RAG document processing with Red Hat OpenShift AI for production-ready parsing, embedding, and querying of PDF documents.

Learn how to develop prompt injection guardrails for NeMo Guardrails on your local machine with this step-by-step guide.

Save on GPU costs with GPU-pruner, an open source tool for Kubernetes that scales down idle GPU workloads and reclaims resources.

Learn how to implement hermetic builds for Open Data Hub and Red Hat OpenShift AI notebook images, ensuring reproducibility, auditability, and compliance.

Learn how AI observability works with MLflow, an open source tool that can trace an agentic query and find the root cause of discrepancies between AI assistant responses and dashboard data. This post discusses the practical value of tracing an agent and what MLflow adds to an observability stack.

This article demonstrates how to use Helm to manage Red Hat OpenShift AI and its dependencies on Red Hat OpenShift, from a full platform installation to a lightweight inference-only deployment.

Learn how to fine-tune Qwen3-4B for correct tool calls using GRPO on Red Hat OpenShift AI.

Learn how to fine-tune large language models on Red Hat OpenShift AI with Ray and Training Hub for optimized SQL generation using the LoRA algorithm.