AI service data in Splunk AI Assistant for SPL examples

Prompt provided by the customer and the response given by the assistant example

The following is an example of a customer prompt and the assistant response:

This image shows an example of how a prompt entered into the assistant by a customer gets collected as shared data.

SPL generated by the assistant is run in search example

The following is an example of if the generated SPL is run in search:

This image shows an example of how the assistant collects any SPL that a customer generates in a chat, if that SPL is then opened in a Splunk search.

Assistant fails to provide results example

The following is an example of if the assistant is unable to provide results from the user prompt:

JSON
{ [-]
   query: Tell me about the following SPL, whether it is valid in every way, and how it can be improved
   response: Sorry, can you try rephrasing? Changing a word or two can generate a better result.
   trace: { [-]
     kinesis: { [-]
       approximateArrivalTimestamp: 1727975013646
       deliveryStreamArn: arn:aws:firehose:us-east-1:671568874969:deliverystream/dataeng-data-sc-splunk
       sequenceNumber: 49655134549526570122125560322351985856941311528865366066
       shardId: shardId-000000000003
       sourceKinesisStreamArn: arn:aws:kinesis:us-east-1:671568874969:stream/dataeng-data-sc
     }
     lambda: { [-]
       invokedFunctionArn: arn:aws:lambda:us-east-1:671568874969:function:dataeng-splunk-firehose-transform-sc:$LATEST
       logStreamName: 2024/10/03/[$LATEST]ff089c4f182f4aa69de2b30dc9127cd0
       timestamp: 1727975074574
     }
     ngx: { [-]
       event_time: 1727975013
       host: telemetry-splkmobile.dataeng.splunk.com
       http_x_forwarded_for: 100.29.68.135,100.67.61.130
       http_x_request_id: 0793723e-1d66-9a6a-972c-095bf9f01649
       proxy_add_x_forwarded_for: 100.29.68.135,100.67.61.130, 127.0.0.6
       request_id: a5868b1596b8039f4152874c7565d815
       timestamp: 1727975013
       version: 0.0.96
     }
     scs: { [-]
       environment: prod
       region: region-iad10
       tenant: 46128259c744b8e01fd735a62500146cdda1d8a7b838a5054fa4f9e7fc3f2ce2
     }
   }
   type: empty_retrieval_QA_LLAMA_INDEX
}

Feedback submitted by user example

The following is an example of feedback submitted by a user:

This image shows an example of how customer feedback gets collected by the assistant.

Elapsed time to complete a request example

The following is an example of the elapsed time to complete a request in the assistant:

CODE
2024-10-08 16:03:34 UTC, Level=INFO, Pid=3351026, Logger=AsyncHttpJobs, File=jobs.py, Line=147, UUID="3319264e-73ae-40fd-a772-3c5ef8e6b109", message="Generation complete.", user="1803185961117335798"
2024-10-08 16:03:34 UTC, Level=INFO, Pid=3351026, Logger=AsyncHttpJobs, File=jobs.py, Line=140, UUID="3319264e-73ae-40fd-a772-3c5ef8e6b109", apply_time="7.87583", user="1803185961117335798"