View Only
  • 1.  Guidance required in implementing Watsonx.ai usecase

    Posted Thu December 28, 2023 05:37 PM


    As a newcomer to Watsonx, I have limited hands-on experience but aim to utilize its capabilities for building a chatbot to address customer complaints. My primary query pertains to integrating Watsonx with my existing database, which contains the data we manually check before responding to complaints. Specifically, I am seeking guidance on automating responses using Watsonx rules to query and respond based on the data in our database.

    For instance, if a bank customer complains about not receiving a postcard activation email, the current manual process involves checking the customer's status in our system databases. I aim to streamline this response by leveraging Watsonx to automate the process of querying databases and responding accordingly.

    Any assistance or guidance on achieving this automation through Watsonx would be immensely appreciated.

    Thank you.


    Shahmeer Arshad

  • 2.  RE: Guidance required in implementing Watsonx.ai usecase

    Posted Fri December 29, 2023 12:32 AM

    These might help you:

    article: https://developer.ibm.com/articles/use-generative-ai-intelligent-workflow-automation-watsonx/
    tutorial: https://developer.ibm.com/tutorials/awb-implement-rag-using-watsonx-platform/


  • 3.  RE: Guidance required in implementing Watsonx.ai usecase

    Community Leadership
    Posted Wed January 03, 2024 11:35 AM

    Hello Shahmeer,

    Happy New Year! I hope that you had an enjoyable holiday.

    I wanted to follow-up with you to see if the documents suggested by Supal were helpful in responding to your question. If you need additional assistance please let me know.

    Thanks very much,


    AI Community Manager

    Nick Plowden
    Community Engagement

  • 4.  RE: Guidance required in implementing Watsonx.ai usecase

    Posted Fri April 26, 2024 02:33 PM

    Hi Shahmeer,

    You can also train the foundation model, by giving input out put pairs. A synthetic data set on What could be the user's input and what should be the ideal corresponding response by the system.  The trained model can assist by identifying what is the customer's message and accordingly perform the responsive action.

    We could connect offline, if you want more guidance.
    Arvin, +91 96000 38297

    Arvin Subramanian