Skip to content

AI Lifecycle on IBM Z

AI Lifecycle Overview

IBM offers a broad set of AI technologies for IBM Z. Many products support more than one stage of the AI lifecycle and often work together. Use the sections below to explore where each product is commonly used, understand how products relate to one another, and see common paths organizations take when building AI solutions.

  • Prepare Data

    Clean, transform, and engineer features to prepare data for AI models.

  • Build Models

    Build traditional ML and deep learning models on IBM Z or hybrid cloud, wherever your investments are.

  • Deploy Models

    Convert, optimize, and deploy your model into a production environment on IBM Z.

  • Serve Model

    Run trained models in production inference environments where they can process requests, generate predictions, and support operational workloads.

  • AI Governance

    Monitor, explain, and govern models in production for bias, drift, and regulatory compliance.

IBM Z Products Used in the AI Lifecycle

Explore which products and tools support each stage of the AI lifecycle across IBM z/OS and Linux on Z.

Product Prepare Data Build Models Deploy Models Serve Model AI Governance IBM z/OS Linux on Z
IBM Synthetic Data Sets YesYes
IBM Z Platform for Apache Spark Yes—
IBM Z Accelerated SnapML* ★ Yes via zCXYes
IBM Z Accelerated for TensorFlow* ★ Yes via zCXYes
IBM Z Accelerated for PyTorch* ★ Yes via zCXYes
Python AI Toolkit for z/OS ★ Yes—
ONNX & IBM Z Deep Learning Compiler* Yes via zCXYes
IBM Z Accelerated Serving for TensorFlow* ★ Yes via zCXYes
IBM Z Accelerated for Triton Inference Server* ★ Yes via zCXYes
Machine Learning for IBM z/OS ★★Explainability & Drift Yes—
AI Optimizer for IBM Z and LinuxONE —Yes
Red Hat OpenShift AI ★TrustyAI Yes via zCXYes
Red Hat AI Inference Server —Yes
watsonx.ai ★ —Yes
watsonx.governance —Yes

* Available in IBM Container Registry, with support included in AI Toolkit for IBM Z and IBM LinuxONE

⭐ = Most used capability