Skip to main content

Getting Started with AI Agent Observability

To start WhaTap AI Agent Observability, create a project, select the agent type that matches your LLM application language, and follow the installation guide to apply the agent. Once the agent is applied, you can view the collected data on the LLM dashboard.

WhaTap AI Agent Observability monitoring starts with the following steps.

Step 1. Sign up​

To use the WhaTap monitoring service, you must first sign up. For more information about signing up, see the Account Management > Sign Up documentation.

Step 2. Create a project​

Create a project before installing the agent.

  1. Log in to the WhaTap Monitoring Service.

  2. In the upper left corner of the screen, click icon All Projects > add icon Create project.

  3. ❶ On the Select product screen, select the AI Observability category.

  4. Click the Create button for the AI Agent Observability product.

  5. ❷ On the Create project screen, enter the project information.

    ItemDescription
    Project nameEnter a name that identifies the project.
    Data server regionSelect the server region where your data is stored. Your data is stored in the selected region.
    Time zoneSet the reference time used for alerts and reports.
    Notification language settingSelect the language for alert messages. Korean and English are supported.
    Project groups (optional)You can classify projects into groups.
    Project description (optional)You can enter an additional description for the project.

  6. After entering all items, click the Creating a project button.

Note

When adding a project with an organization selected, you must set the Groups of organization. For more information about groups, see the following documentation.

Step 3. Select an agent type​

After creating a project, you are automatically redirected to the ❸ Agent installation page. Select the agent type that matches your LLM application language. After selecting an agent type, you are redirected to the installation guide page for that language.

Agent type selection screen

Agent TypeDescriptionStatus
PythonAgent for Python-based LLM applicationsSupported
JavaAgent for Java-based LLM applicationsComing soon
Node.jsAgent for Node.js-based LLM applicationsComing soon
OthersLLM applications based on other languagesComing soon
Note

If you are not redirected to the Agent installation page, select Management > Agent installation from the left menu.

Step 4. Check the access key​

The access key is a unique ID for activating the WhaTap service. In the first step of the agent installation guide page, click the Getting the access key button to issue the access key.

You must enter the issued access key in the agent configuration file (whatap.conf).

Step 5. Install the agent​

Install the agent according to the agent type you selected. For detailed installation instructions for each agent type, see the following documentation.

Step 6. Check the dashboard​

After you install the agent and restart the application, data starts being sent to the WhaTap collection server within a few minutes.

  1. In the left sidebar, click Dashboard > LLM Dashboard.

  2. When LLM API calls occur, the request count, token usage, cost, and performance metrics appear in the widgets.

Last updated on