Red Hat Expands Enterprise AI for Hybrid Cloud

Red Hat, Inc., the world’s leading provider of open-source solutions, has announced the latest enhancements to Red Hat AI, a comprehensive suite of products and services designed to accelerate AI development and deployment across the hybrid cloud. Red Hat AI offers an enterprise-grade AI platform optimized for model training and inference, delivering improved efficiency, a seamless user experience, and the flexibility to deploy across any hybrid cloud environment.

As businesses strive to reduce the costs associated with deploying large language models (LLMs) at scale for various use cases, they face challenges in integrating these models with proprietary data while ensuring accessibility across data centers, public clouds, and edge environments.

Red Hat AI, which includes Red Hat OpenShift AI and Red Hat Enterprise Linux AI (RHEL AI), addresses these challenges by providing an enterprise AI platform that supports the adoption of optimized models tailored to business-specific data. This enables seamless deployment across hybrid cloud environments for both training and inference, leveraging a wide range of accelerated computing architectures.

Red Hat OpenShift AI

Red Hat OpenShift AI serves as a complete AI platform for managing predictive and generative AI (Gen AI) lifecycles within the hybrid cloud. It includes machine learning operations (MLOps) and LLMOps capabilities, offering tools for predictive model development, Gen AI tuning, and streamlined model management, covering data science pipelines, model monitoring, governance, and more.

The latest release, Red Hat OpenShift AI 2.18, introduces several key updates aimed at optimizing AI models and enhancing efficiency across hybrid cloud environments. Notable features include:

• Distributed Serving: Through the vLLM inference server, distributed serving allows IT teams to distribute model serving across multiple GPUs. This optimizes computing resources, reduces the load on individual servers, accelerates training and fine-tuning, and enhances service distribution across AI model nodes.
• End-to-End Model Tuning Experience: Leveraging InstructLab and Red Hat OpenShift AI data science pipelines, this feature streamlines the fine-tuning process for LLMs, improving scalability, efficiency, and auditability in large-scale production environments. Users can manage these processes through the Red Hat OpenShift AI dashboard.
• AI Guardrails: Available as a technology preview, AI Guardrails enhances LLM accuracy, performance, latency, and transparency. It helps IT teams detect and mitigate issues such as hateful or abusive language, personally identifiable information, competitive intelligence leaks, and data restricted by corporate policies.
• Model Evaluation: Utilizing the language model evaluation (lm-eval) component, this feature provides data scientists with critical insights into model performance across various tasks, including logical reasoning, mathematical analysis, and adversarial natural language processing. This ensures the development of more effective and responsive AI models.

Also Read: Voxel51’s New Workflows Speed Up Visual AI for Enterprise

Red Hat Enterprise Linux AI (RHEL AI)

RHEL AI, a key component of the Red Hat AI portfolio, provides a foundation model platform for consistent development, testing, and execution of LLMs in enterprise applications. RHEL AI integrates Granite LLMs and InstructLab model alignment tools within a bootable Red Hat Enterprise Linux server image, facilitating deployment across hybrid cloud environments.

With the launch of RHEL 1.4 in February 2025, the platform introduced several enhancements, including:

• Granite 3.1 8B Model Support: The latest addition to the open-source-licensed Granite model family brings multilingual capabilities for inference and taxonomy/knowledge customization (developer preview), along with a 128k context window for enhanced summarization and retrieval-augmented generation (RAG) tasks.
• Graphical User Interface for Skills and Knowledge Contributions: Available as a developer preview, this feature simplifies data ingestion and chunking while enabling users to contribute skills and knowledge to AI models.
• Document Knowledge-Bench (DK-Bench): This tool facilitates the comparison of fine-tuned AI models using private data against their base versions, ensuring optimized performance and relevance.

Red Hat AI InstructLab on IBM Cloud

Enterprises are increasingly prioritizing AI solutions that balance accuracy, data security, cost efficiency, and simplicity. Red Hat AI InstructLab, deployed as a service on IBM Cloud, is designed to streamline AI model training and deployment while improving security and scalability. By simplifying the InstructLab model tuning process, businesses can create more efficient AI models tailored to their specific needs while maintaining control over their data.

No-Cost AI Foundations Training

AI is reshaping the way businesses operate and compete. To support organizations navigating this transformation, Red Hat now offers AI Foundations online training courses at no cost. These courses provide AI learning certificates suitable for both senior leaders and AI novices, offering essential knowledge on AI’s role in business transformation, decision-making, and innovation. The AI Foundations training also guides users in applying these concepts with Red Hat AI solutions.