Optimized Inferencing and Integration with AI Platforms on IBM Z

Companies continue to invest heavily in ways to leverage Al to improve their business and differentiate themselves. For IBM Z clients, leveraging Al is also a key aspect of modernizing their existing business critical applications while maintaining strict SLAs and security requirements. The goal of this residency is to provide a Redpaper that dissects a specific use case to introduce and explain AI technologies on IBM Z. It will describe and demonstrate the implementation and integration of an end-to-end solution, including a z/OS CICS application, and discuss the benefits derived from the solution. This includes the different options clients have to do their Al model and development training anywhere and optimally deploy on IBM Z for inferencing. It will also include recommended considerations and best practices around choosing a specific approach, The Redpaper will also discuss how clients can perform model development and training on the cloud and integrate to deploy for inferencing on Z. Technologies to be covered include, but are not limited to TensorFlow Serving, IBM Watson Machine Learning for z/OS (WMLz), IBM Cloud Pak for Data (CP4D), z/OS CICS application, Open Neural Network Exchange (ONNX) and Deep Learning Compiler.