Document Automation - Unstructured Data Learning Instance AARI Request Failed

​@Manik.Sharma31 The only difference in those document types is the use of AI. Check with your IT department to see if the use of AI is allowed for document extraction.

Hi ​@Aaron.Gleason ,

Since this is being executed within the Automation Anywhere Control Room and I believe Document Automation inherently uses AI for extraction, could you please advise what exactly needs to be validated from the local IT department’s side?

Specifically, we would like to understand:

• What AI-related permissions, policies, or services need to be enabled for Document Automation to work with Unstructured Data document types

• Whether there are any network, firewall, proxy, or security restrictions that could block AI-based extraction in Automation Anywhere

• How we can confirm if AI usage is allowed and properly configured for Automation Anywhere in our environment (Control Room level)

Any guidance on what the local IT team should check or enable would be very helpful. Or should I raise a support ticket with Automation Anywhere?

​@Manik.Sharma31 Is this an on-prem control room or a cloud control room? This question is pivotal.

If it is on-prem, your IT department should easily be able to tell what was blocked. They probably already know if they’re paying attention to the firewall logs.

If it is cloud, we have to look at this a completely different way.

​@Aaron.Gleason It is a cloud control room.

​@Manik.Sharma31 Using Wireshark, it appears there may be some calls to *.algolia.net to do some of the generative AI queries, possibly called from the Bot Agent.

Your IT department can check their firewall logs for any blocked DNS queries or sites, too.