Generate recommendation or follow up questions using copilot studio

(0) Share

Report

Posted on by NaveenR

Hi team,

Could you please help in this

we would like to generate recommendation/suggestions or follow up question based on the response/previous user input/query dynamically.

Example:

We have knowledge document contains product information

if user ask let's say what is product ABC, currently we are getting only response to that product ABC. but we would like to show some follow-up questions along with that like try asking: ABC Price, ABC benefits etc. dynamically not using any manual adaptive card

Thanks

Categories:

General topics

I have the same question (0)

All responses (2)

Answers (0)

Suggested answer

SaiRT14 1,990 Super User 2025 Season 2 on at

Like (0)

Report
i tried below approach:

Power Automate Contextual APIs

Build a Contextual Model: Use a Dataverse table or similar data source to store: Products and their attributes (e.g., price, benefits, usage). Predefined follow-up intents for each product.

Dynamic Trigger: When a user asks about Product ABC, trigger a flow in Power Automate that: Matches the product with the Dataverse table. Retrieves associated attributes and suggested intents.

Generate Recommendations Dynamically: Based on the user’s query, use AI Builder or custom logic to generate personalized suggestions.

let me know if you need more details.

Was this reply helpful? Yes No
NaveenR 20 on at

Like (0)

Report

Thanks for the response @SaiRT14

Here we don't have any product information with us so we can't create any Dataverse tables, and we are calling the API to get the answer for user query.

So, what would be the approach to get follow-up questions/ recommendations?

Is it using AI builder or is there any way?

Example Scenario:

User query: What is product ABC?

Answer: Get the response by calling the API and along with this we would like to show like Try asking: Product ABC price? Product ABC dimensions? Product ABC benefits? etc.

It should generate relevant follow-up questions based on user query.

Was this reply helpful? Yes No