Analyzing Executions with Automic Automation's Gen AI

Automic Automation's Gen AI analyzes and explains the output of your automated processes, it identifies existing and potential issues and suggests solutions to those issues. You can leverage its capabilities to analyze executions from any place in AWI where a runID (that is, an execution) is available.

To Analyze an Execution

  1. Right-click the execution/runID, task, Client, Agent, etc. in one of the following areas in AWI:

    • List of Tasks

      Right-click a task and select Executions > Analyze Execution.

    • Any Executions list

      Right-click and execution and select Analyze Execution.

    • List of Clients

      Right-click a Client and select Monitoring > Analyze Last Execution.

    • List of Agents

      Right-click an Agent and select Monitoring > Analyze Last Execution.

    • Server Processes

      Right-click a process and select Monitoring > Analyze Last Execution.

    • List of Queues

      Right-click a Queue and select Monitoring > Analyze Last Execution.

    • Any report that contains a runID

      Right-click the runID and select Analyze Execution.

    • Global Search

      Right-click an object or task in the Global Search dropdown list and select Analyze Execution (for tasks) or Monitoring > Analyze Last Execution (for objects).

    • Explorer in the Process Assembly perspective

      Right-click the executable object and select Monitoring > Analyze Last Execution.

  2. The Automation AI Assistant panel opens on the right-hand side of the screen.

    Automic Automation's Gen AI crawls all the reports and logs available for the selected execution. It summarizes and analyzes the information and presents it in this panel. This information includes the status and execution key data for successful executions. In the case of failed executions, it also explains why it failed pointing out the root-cause of the errors and suggesting potential fixes. It contains the following sections:

    • Analysis of RunID <runID>, with the details about the run (status, name of the object, execution end time). The runID is a link to the execution. The name of the object is a link that opens it in the Process Assembly perspective and lets you scrutinize its properties, provided you have the necessary rights.

    • Last Change, which indicates User and timestamp of the last change in the object configuration.

    • Reason for Failure, available if the task ended in a failed status. It explains what caused the failure and the error messages.

    • Report section that depends on the type and on the status of the task. It summarizes and explains the content of the report.

    • Open Report link.

  3. You can start a conversation in the context of the execution in the Ask Automation AI Assistant field at the bottom of the pane. You can type any question you need to drill down into the reasons that caused the failure, to request better error handling if you need more clarity as to why it happened, ask for suggestions to better configure the underlying object, and so on. The Automation AI Assistant recognizes and remembers the context in which you ask your questions.

    The Automation AI Assistant responds in natural language. You can ask it to provide the answers in any format that you need (json, markup language, etc.) and you can have follow-up conversations with it in any language, not just in the language of your interface. The more precise that you word your question, the better the answer will be. You can ask your questions in any language.

  4. To clear or exit the chat, click on the three dots to the right of the chat box and select Clear Chat / Exit Chat.

See also: