Red Hat at IBM TechXchange 2025

Data scientists and machine learning (ML) engineers face the challenge of efficiently and cost-effectively deploying large language models (LLMs) for real-time inference. In this session, attendees will learn how vLLM and its specialized optimizations—such as continuous batching, KV-caching, and speculative decoding—can maximize inference throughput while minimizing latency. We’ll introduce Red Hat AI Inference Server a container-native solution that simplifies deploying and scaling LLM inference across hybrid cloud environments. This latest addition to Red Hat AI is available as a standalone or as part of Red Hat OpenShift AI or RHEL AI, delivers an enterprise-supported version of vLLM with integrated model compression features—enabling faster, more cost-efficient inference while helping reduce GPU footprint and operational complexity. Attendees will walk away with hands-on guidance for deploying optimized inference pipelines using vLLM and Red Hat AI in real-world environments.

Today’s conversations about AI are all about LLMs (large language models like ChatGPT) and about creating predictive models to assist decision making. But an AI application is more than that. Designing an AI application has challenges that a traditional application does not. Application architecture for AI poses different challenges to architects and developers. Its anatomy is different. Do we need to manage APIs in AI applications in a different way than we would in a traditional application? How do we take advantage of Agentic AI to integrate intelligence throughout our apps? What is the right architecture to evolve the application logic, as well as changes in the LLM over time?

Platform engineering has become essential in delivering scalable, reliable, and consistent developer experiences. Red Hat Ansible Automation Platform stands out as a powerful tool for platform engineers, offering built-in capabilities to simplify workflow automation for both automation creators and users. In this session, we’ll explore how Ansible Automation Platform helps platform engineers create robust, reusable workflows that streamline complex processes while maintaining flexibility and scalability. We’ll dive into real-world examples where Ansible Automation Platform empowers platform engineers to accelerate automation adoption, bridge the gap between automation creators and users, and build automation workflows that improve operational efficiency. By combining centralized governance, reusable automation content, and integration with popular platform engineering tools, Ansible Automation Platform positions itself as the tool of choice for platform engineers aiming to deliver faster and more reliable automation workflows.Join us and see through technical demos how Ansible Automation Platform can transform your platform engineering practices and unlock new possibilities for your organization.

Are you ready to revolutionize your Java application with the latest AI trend? Let's dive into the world of AI Agents and discover how they can unleash new capabilities! We'll start by showing you how to create and deploy AI Agents in a snap, and put them to work on various tasks. But that's not all - we'll also give them the tools and resources they need to make autonomous decisions and deliver quality results. But, beware: a single agent can quickly become a bloated superhero with too many responsibilities, leading to slower performance, higher costs, and mediocre results. To avoid this, we'll introduce you to the power of multiple, specialized agents that can team up and work together in harmony using Agentic workflow patterns. By the end of this journey, you'll have created a cutting-edge, enterprise-ready Java application that's powered by versatile AI Agents. Are you ready to unleash their potential?

Organizations are increasingly turning to AI for transformative business insights and task automation. This session offers a practical introduction to LLM-as-a-Service (LLMaaS), showing you how to build and manage custom Large Language Models (LLMs) to meet specific enterprise needs. These LLMs can be served for business insights as well as linked to other AI agents to perceive the world around them and complete tasks. We'll cover deploying LLMs securely within your private cloud, ensuring data security, scalability, and compliance. You'll get a clear overview of integrated services, access controls, and usage metrics for chargeback. We'll also explore practical, high-impact use cases like AI-powered code assistants for developers and agentic AI applications, demonstrating the immediate business value LLMaaS can deliver.

Everyone wants to have artificial intelligence (AI)-enabled applications that are private, scalable, and security-focused. However, this is often complex, as you must either bring together the needed services and tools yourself or use public cloud services that require data to be uploaded or shared. In this session, you will learn how to use a service architecture model with Red Hat OpenShift AI and other tools from Red Hat's portfolio. With this approach, IT teams can become internal, private AI providers "à la OpenAI"—giving their internal customers access to the tools they really need, whether they are developers or not.

Enterprise development teams face significant hurdles that impede productivity and innovation, directly impacting business outcomes. Challenges such as convoluted onboarding processes, reliance on tribal knowledge for application spin-up, CI/CD pipeline choke points, disconnects in security and compliance, fragmented AI/ML development and operations, and modernization roadblocks lead to delayed project timelines, frustrated development teams, and missed market opportunities. These inefficiencies translate into tangible costs from wasted developer time and increased operational risk. This session explore strategies and solutions to re-imagine the developer experience (DevEx). We will focus on the principles and implementation of Platform Engineering to create a unified, self-service, and automated developer journey. We will also discuss how DevEx can transform an organization's ability to deliver value, accelerate innovation, and foster a more productive and engaged development workforce, ultimately driving business success.

Enterprises seek Agentic AI for autonomous automation, yet struggle with complex workflow orchestration, ensuring reliable behavior, and efficiently scaling deployments. This session will walk you through how Red Hat AI can help address these critical challenges through an open, hybrid platform that simplifies Agentic AI. We'll detail how our unified APIs, leveraging technologies like MCP (Model Context Protocol) and Llama Stack, streamline agent workflow assembly and enable governed, trustworthy deployment. You’ll also hear how our scalable infrastructure ensures cost-optimized operations. A practical demonstration will highlight an Agentic AI solution proactively managing at-risk customer renewals by intelligently integrating with a CRM (Customer Relationship Management), utilizing RAG, and automating communication via Slack, revealing how Red Hat AI turns complex problems into tangible business value.

Get started with RHEL 10 today! After a short briefing on RHEL 10 updates, you will get hands-on with several of the core new RHEL 10 features like: Building and managing image mode built images Working with the command-line assistant, an on-system AI driven assistant to help you manage your systems Implementing security standards with RHEL system roles, powered by Ansible. Of course, with RHEL 10 machines at your disposal, you can also explore the latest RHEL distribution with the topics interested to you as well!

As application development paradigms evolve, Linux has emerged as a preferred environment for developers, particularly in enterprise and production contexts. For years, Windows-based developers have relied on virtual machines (VMs) to emulate Linux. And although dependence on full VMs has diminished with the rise of container technology, there is still a need for VM layers on non-Linux operating systems. Red Hat Enterprise Linux (RHEL) 10 offers developers an alternative to a container-based or layered VM setup. In this session, we’ll share how this robust development environment with modern tools and packages positions Linux as a no-overhead platform for development— We’ll provide an overview of the latest developer-oriented RHEL features that help improve workflows for efficient application development, making Linux the default platform for serious developers.

In this interactive lab attendees will learn how Red Hat Ansible Automation Platform, large language models (deployed on Red Hat Enterprise Linux AI or Red Hat OpenShift AI), Red Hat Ansible Lightspeed, and forthcoming policy-as-code capabilities integrate to enrich operational data, intelligently generate playbooks, and maintain governance in different IT scenarios. Anshul Behl, Sean Cavanaugh, and Marty Turner will cover AI-Ops principles and demonstrate how flexible, closed-loop workflows can drive transformative outcomes. Attendees will leave with a roadmap for how to use AI and the Ansible Automation Platform to accelerate automation and innovation.