Gen AI on IBM Z

AI and Automation on IBM Z

AI and Automation on IBM Z

AI and Automation on IBM Z is a one-stop destination to find the latest communications, news, videos and events, as well as collaborate and network.


#DevOps
#IBMZOS
#IBMZ
#DevOpsSolutions
#GenerativeAI

 View Only

  • 1.  Integrating Watson/watsonx/Project Bob on IBM Z for COBOL Code Analysis, Improvements & Performance Optimization

    Posted 07/07/26 02:57 AM
    Edited by Dennis Köcher 07/07/26 02:57 AM

    Hello everyone,

    I am a COBOL and IBM Z Mainframe developer. I want to use AI (especially Watson / watsonx) directly on the mainframe for:

    • Analysis and understanding of large, legacy COBOL codebases
    • Code improvement and refactoring suggestions
    • Performance analysis and optimization support
    • General static and dynamic code analysis

    I am looking for practical experiences on how to integrate such AI capabilities on IBM Z.
    Specific questions:

    • How do you set up Watson Machine Learning/watsonx/Project Bob on IBM Z in a real project?
    • Which architectures work best (z/OS Container Extensions, IZOA, Python on Z, direct integration, etc.)?
    • How do you effectively connect AI models to existing COBOL applications and codebases?
    • What tools and best practices exist specifically for code analysis and optimization on the mainframe?
    • Lessons learned: What works well? What was complex or disappointing?

    I'm particularly interested in technical, hands-on implementations rather than general marketing material.

    Any tips, whitepapers, blogs or real-world experiences would be greatly appreciated.


    Thank you very much!


    Best regards,

    Dennis



  • 2.  RE: Integrating Watson/watsonx/Project Bob on IBM Z for COBOL Code Analysis, Improvements & Performance Optimization

    Posted 23 days ago

    Hi Dennis Köcher

    If your goal is AI-assisted COBOL understanding and modernization, I'd recommend starting with watsonx Code Assistant for Z rather than building everything from scratch. It can help analyze legacy COBOL, explain code, and support modernization workflows. For production environments, keeping AI services separate from core transaction processing (via APIs or z/OS Connect) tends to be more practical than embedding AI directly into COBOL applications. I'd also be interested to hear real-world experiences with watsonx Code Assistant for Z, especially around large-scale code analysis and performance optimization.



    ------------------------------
    Elena Mia
    Technical Consultant
    Damco Solutions
    plainsboro NJ
    6096320350
    ------------------------------



  • 3.  RE: Integrating Watson/watsonx/Project Bob on IBM Z for COBOL Code Analysis, Improvements & Performance Optimization

    Posted 14 days ago

    Hi Dennis,

    It's great to see interest in applying AI to Z application analysis and modernization. This is an area that is evolving rapidly, and several solutions are now available that can help tackle challenges such as understanding large COBOL codebases, dependency analysis, refactoring, and modernization.

    Based on the use cases you described, I would recommend taking a look at Bob Premium Package for Z, which is the evolution of watsonx Code Assistant for Z.

    Useful resources:

    For COBOL application analysis, this is probably the most relevant IBM offering today. It combines AI capabilities with static analysis and IBM Z domain knowledge, allowing it to understand applications in the context of z/OS technologies, middleware, dependencies, and more capabilities for Understanding large COBOL codebases, Application and dependency discovery, Impact analysis, Refactoring support etc.

    If you're interested in evaluating Bob Premium Package for Z, I'd suggest contacting your IBM representative to discuss trial availability and adoption options.

    For AI and Machine Learning workloads on IBM Z more broadly, you may also find the following resources useful:

    The trial includes three days of hands-on labs and is a good opportunity to explore AI and Machine Learning use cases on IBM Z in a practical environment.

    Hope this helps and have fun exploring these technologies!

    Eva



    ------------------------------
    Eva Anahi Berrio
    ------------------------------



Related Content