GAPTEQ Assist can be used to check database content for specific topics / keywords - it provides direct feedback or executes suitable actions.
Note! Using the demo example “Product management furniture store Müller”, we will now describe how existing database content can be automatically checked or queried and how the feedback can be automatically processed (e.g. added to other input fields).
Initial situation
In this GAPTEQ application of a furniture store, all product information (name, short description, detailed product description, dimensions, supplier, ...) is recorded in a form. Each product belongs to a product category and is initially recorded with a product name and a brief description.
Aim
GAPTEQ Assist should now be used to check whether relevant information about the product (e.g. dimensions) is already recorded in the database - if so, this should be transferred to the short description of the product.
1. Place AI chat component & define triggering event
- To do this, the AI chat component must first be dragged onto the page (see the screenshot above in the bottom right-hand corner).
By clicking on a button (“Add short description”), the short description of the item should now be supplemented with the specific dimensions of the item.
- The action "ask(question, showQuestion)" can now be stored for the "clicked" event of the button - referenced to the AI chat component.
- A string expression with the corresponding prompt to the AI is stored for the "question" parameter. ("Check whether this information is included in the text and add it to the end of the text if it is not already included")
- By specifying the corresponding form fields (here: "formProduct.row.Dimensions"), the AI knows which form fields are to be checked for content (see screenshot at 2.).
2. Define data sources for Promt
- The “selected.Row” command on the “ShortDescription” form field gives the AI the corresponding data source that is to be translated in this case.
- The “showQuestion” parameter is set to “True” with a Boolean.
3. Transfer generated answer to desired form field
The checked information should now be transferred directly to the form field, i.e. without having to copy and paste the generated text from the AI chat component.
- To do this, you can use “Set a value” on the desired form field (here: “form1_ShortDescription”) to accept the generated response directly.
Note! The AI chat component is also available as an invisible component, so that the chat answers provided by the AI are not openly displayed in a chat on the page, but can be used directly.