The exponential growth in data over the last decade, coupled with adoption of artificial intelligence, has opened up organizations to valuable insights, opportunities for improvements in efficiency, and increases in profits. Organizations can deliver more business value both internally to employees and shareholders and externally to clients and coInstead, businesses embrace the strategy of data gravity for AI. Transaction gravity is the force that alludes to the high number of transactions that move through IBM Z. For organizations to get real-time business and operational insights at scale, the AI workloads must be colocated with the transactional applications and where the data originates.

Machine Learning for IBM z/OS is an IBM offering that helps organizations deploy machine learning models and infuse them with AI on IBM z/OS environments natively that are close to their most mission-critical workloads. In the secure Z environment, your machine-learning scoring services can coexist with your transactional applications and data. This configuration helps to support high throughput, minimize response time, and deliver consistent service level agreements (SLAs).

This book introduces Machine Learning for IBM z/OS version 3.2.0 (MLz) and describes its unique value proposition in addition to the business value of AI on the IBM Z platform. It provides guidance for you to get started with aligning your business goals and use cases for AI. It outlines the various patterns for accessing and using IBM Z data on-premises, in the cloud, and in hybrid environments. It discusses in detail the steps and best practices for model building, deployment, inferencing, and post deployment monitoring.

The book includes examples of how you can use the MLz web-based user interface to train, evaluate, and deploy a model. The book explores the value of MLz by using a batch job elapsed time prediction example and shows how to apply the batch job elapsed time model with MLz for application integration. The book provides insight into how MLz can be used within machine learning operations to run AI on IBM Z. This book also examines use cases across industries to illustrate how you can turn your data into valuable insights with MLz.nsumers. The data that fuels these insights is the foundation that these organizations must use to innovate. Organizations must capitalize on advanced AI capabilities to stay ahead of the competition and operate in a secure environment that protects data throughout its lifecycle and enables real-time access.

IBM Z can help provide a secure environment. Its multi-workload transactional platform powers the core business processes of many Fortune 500 enterprises. The IBM Z platform can provide security, availability, reliability, and scalability. With core transactions and data originating on IBM Z, it makes sense for AI to exist and run on the same platform.

In the past, some businesses chose to move their sensitive data off IBM Z. However, some businesses have reconsidered their decision because of the massive growth of digital data, the punishing cost of security exposures, and the unprecedented demand for real-time actionable intelligence from data.

Chapter 1. Unlocking the business value of AI on IBM Z

Chapter 2. Building business-driven and strategic use cases with Machine Learning for IBM z/OS

Chapter 3. The art of data engineering

Chapter 4. Model building for IBM Z

Chapter 5. Model deployment and inferencing

Chapter 6. Streamlining AI from models to applications by using machine learning operations