Introduction to the AI Framework for z/OS
IBM® z/OS® 3.1 introduces artificial intelligence (AI) that is infused into the operating system to provide intelligent automation and accelerated inferencing at scale to extract and leverage valuable data insights. The AI capability in z/OS consists of the following deliverables:
- AI Framework for IBM z/OS
- AI Framework for IBM z/OS (z/OS AI Framework) supports augmenting z/OS with intelligence that helps to optimize IT processes, simplify management, reduce skill requirements, and improve performance. AI Framework for z/OS includes components that provide data collection, AI model training, deployment, and monitoring, and AI inferencing and scoring. Further, AI Framework for z/OS provides a component that allows other z/OS components to plug into the framework by enabling traditional z/OS programming interfaces, such as assembler services, to access REST API functions for simpler AI deployment.
- AI System Services for IBM z/OS
- AI System Services for IBM
z/OS (z/OS AI
Services) delivers foundational AI capabilities
and represents one of the key components of the z/OS AI
Framework. AI System Services for z/OS provides the following capabilities:
- IT data ingestion and filtering capabilities that enable the collection of data for model training and inference purposes
- AI model server capabilities that support the AI model lifecycle, including AI model training, deployment, inferencing, monitoring, and retraining
The integration of AI System Services for z/OS with the rest of the AI framework components enable prebuilt AI models to be put into operation to simplify the management of z/OS and its offerings by augmenting them with AI, reducing skill requirements, optimizing IT processes, and improving performance.
- AI-powered WLM batch initiator management
- Workload management (WLM) is the first z/OS component to plug into the z/OS AI Framework as a provider of an AI-infused use case. AI-powered WLM batch initiator management is designed to intelligently predict upcoming batch workload and proactively optimize system resources to meet the demand.
The AI Framework for z/OS, including AI System Services for z/OS, offers an AI platform that seamlessly integrates a set of components to enable the operation and use of AI-infused capabilities by z/OS base components, starting with the initial use case for AI-powered WLM batch initiator management. The platform is intended to offer simplified installation, setup, and management of AI-infused capabilities without requiring additional data science or AI skills. The platform is designed to pave the way for AI use case providers to harness the foundational AI capabilities and accelerate the deployment of future AI use cases.