All Products
Search

This Product

    STAROps:Intelligent Conversation quick start

    Document Center

    STAROps:Intelligent Conversation quick start

    Last Updated:Jun 23, 2026

    Run your first natural language data query with STAROps Intelligent Conversation and experience AI-powered conversational O&M.

    What you'll accomplish

    After this quick start, you will be able to:

    • Start an Intelligent Conversation from the STAROps console.

    • Ask a Digital Employee questions in natural language and retrieve observability data query results.

    Before you begin

    • A Digital Employee has been created.

    • The Workspace is connected to the observability data of your target system.

    Step 1: Open the Intelligent Conversation interface

    1. Log on to the STAROps console.

    2. In the sidebar, click New Conversation to open the Intelligent Conversation interface.

    Step 2: Select a Digital Employee

    At the top of the conversation interface, select a Digital Employee. Each Digital Employee has its own capability configuration and data access scope.

    Note

    If only one Digital Employee exists, the system selects it automatically.

    Step 3: Start a conversation

    In the conversation input box, describe your query in natural language. The following table lists common query examples:

    Scenario

    Example query

    Data query

    Show the top 5 hosts by CPU utilization over the past 24 hours.

    Anomaly detection

    What alerts have been triggered in the last hour?

    Root cause analysis

    Analyze the cause of high response latency for service-payment.

    After you enter your question, press Enter or click the send button. The Digital Employee processes the query and returns results.

    Step 4: Review the results

    The Digital Employee invokes the relevant tools to query data and presents results as text summaries, data tables, or metric trend charts.

    To dig deeper, continue the conversation with follow-up questions. For example:

    • "What is the root cause of this alert?"

    • "Compare the data with the same period last week."

    Learn more

    Now that you know the basics, explore these additional capabilities:

    • @ entity reference: In the conversation input box, type @ to reference entities in the Workspace (such as Services, Nodes, and Pods), helping the AI precisely locate the analysis target. For details, see CloudMonitor Intelligent Assistant.

    • / command: The Digital Employee supports the / command to invoke capabilities such as /Skill, /Tool, /clear, and /compact.

    • Long-running tasks: For scheduled inspections or automated O&M, see the Long-running Task Quick Start.

    • Configure default rules: Default rules for a Digital Employee significantly improve conversation quality. For details, see Configure Default Rules.

    FAQ

    Operation failed due to insufficient permissions

    Check the following possible causes:

    • Verify that the Digital Employee has the required permissions.

    • Check whether observability data is connected to the Workspace. If the data source configuration is incomplete, the Digital Employee may be unable to process query requests.

    • Verify that the RAM role associated with the Digital Employee has read permissions for the corresponding data source.

    The Digital Employee returns inaccurate results

    Try the following approaches to improve answer accuracy:

    • Be specific when asking questions. Include context such as the time range, target resource, and metric name. For example, change "check CPU" to "show the top 5 pods by CPU utilization in prod-cluster over the past 24 hours".

    • Use the @ entity reference feature to precisely specify the analysis target. Type @ in the input box, select the target entity, and then send the message.

    • Configure default rules for the Digital Employee to define output format and analysis depth requirements. For details, see Configure Default Rules.