IBM has launched new agentic AI capabilities in IBM Z to elevate productivity, simplify operations, and enhance the user experience.
“With AI foundational services that support security integration, scalable performance and real business impact, organisations can bring AI into their core systems to unlock new levels of productivity and insight,” wrote Skyla Loomis, general manager, IBM Z Software, in an announcement.
New agentic AI capabilities
The new agentic AI framework powers tools such as IBM watsonx Assistant for Z, which enables context-aware responses and supports effective incident remediation.
IBM is also transforming how developers modernise mainframe applications through IBM watsonx Code Assistant for Z by leveraging AI and automation.
To support enterprise-grade AI, IBM Z is evolving with a set of foundational services designed to make AI integration secure, scalable, and impactful. These services form the backbone of the IBM Z AI ecosystem, working with existing infrastructure to:
- Simplify how AI accesses data and executes commands across z/OS and middleware such as Db2, CICS, IMS, and MQ through the Model Context Protocol (MCP), which provides a consistent integration layer for AI agents.
- Support a range of Large Language Models (LLMs), including IBM’s Granite models, and third-party and open-source options from Meta and Mistral.
- Include an Agent Builder within watsonx Assistant for Z, enabling the creation of agents that plan, reason, and act.
- Offer the IBM AI Optimiser for Z, a tool that helps organisations deploy, integrate, and monitor AI solutions with reduced time, cost, and complexity.
Building on the momentum of the IBM z17 launch earlier this year, IBM also unveiled the IBM Spyre Accelerator for IBM z17, designed to work seamlessly with the advanced IBM Telum II processor, enhancing AI performance and enabling the mainframe to run sophisticated AI models.
