Forum Discussion

msprog's avatar
msprog
Advocate III
5 months ago
Solved

facing issue in the MS tutorial for data agent on Graph model

Hey 

I am new to Fabric IQ and was trying out the tutorial on ms learn.

 

https://learn.microsoft.com/en-us/fabric/iq/ontology/tutorial-4-create-data-agent

on step 4, i have created a data agent.

When i prompt the agent, i get error: please see below.

 

 

{"entitySelector":{"query":"MATCH (store:Store)-[:has]->(sale:SaleEvent)<-[:soldIn]-(product:Products)\nRETURN store.StoreId AS store_id, product.ProductId AS product_id, store.StoreName AS store_name, product.ProductName AS product_name","queryType":"GQL"}}

 

Please can anyone explain what is going wrong here?

 

thanks

 

 

 

 

  • Great msprog , then we should review the query and analyze why the model is not interpreting it correctly.

    As a final step, if needed, we could provide additional instructions to the agent.

    If you were working with a Lakehouse, you could also include example questions with queries in the agent, so that if something similar comes up, it can take it into account during interpretation.

4 Replies

  • Hi msprog 
    Quick question: even if you're following the Fabric IQ and agents tutorial, are you using a paid Fabric capacity or a trial one?
    It won’t work with a trial capacity. In fact, the agent prerequisites specify that a paid capacity is required.
    Also, please check in the admin portal that Fabric Agents and Graph features are enabled.
    Hope this helps.
    Let me know if you’ve managed to solve it or if you’re still facing issues.
    Best regards,

    • msprog's avatar
      msprog
      Advocate III

      arabalca  it is only this prompt that is failing - i suspect something wrong with the query the agent has constructed. 

      other prompts worked for e.g. below 

       

      • arabalca's avatar
        arabalca
        Super User

        Great msprog , then we should review the query and analyze why the model is not interpreting it correctly.

        As a final step, if needed, we could provide additional instructions to the agent.

        If you were working with a Lakehouse, you could also include example questions with queries in the agent, so that if something similar comes up, it can take it into account during interpretation.

  • Hi msprog ,

    if this helped you, I’d appreciate it if you could mark it as the accepted solution and give it a like 👍

    This helps others in the community find the answer more easily.

    Thanks! 🙌