Skip to main content

AI Search agent

What is AI Search agent?​

The AI Search agent is a powerful tool designed to search and retrieve information from various connected data sources (AI Connectors) using natural language queries. It acts as a central interface for querying information across different platforms that your organization uses.

The AI Search agent provides the following capabilities:

  • Unified search: It can search across multiple types of AI Connectors, including files, Confluence, web pages, and other platforms.

  • Source filtering: You can apply filters to specify which connected sources the agent should search within for your query. This allows you to narrow down the search to relevant platforms like Jira, web connectors, or file connectors.

  • Natural language queries: Interact with the agent by typing your questions in plain language.

  • Contextual answers: The agent provides answers by retrieving relevant information from the indexed data within your selected AI Connectors.

  • Source attribution: For each answer, the agent indicates the source document(s) from which the information was retrieved, along with a match percentage, helping you verify the information.1

Configuring the AI search agent​

The AI Search agent's functionality relies on the configuration of AI Connectors. These connectors link the agent to your data sources.

  1. From the DataGOL Home page navigate to the Connectors page.

  2. Click on the New Connector button.

  3. Select connector type: Choose the type of platform you want to connect (e.g., File, Confluence, Web). The system will display the currently supported connectors.

  4. Enter configuration details: Depending on the connector type, you'll need to provide specific information:

    • For a File connector: Upload the file you want the AI to search through.

    • For a Confluence connector:

      • Name of connector: A descriptive name for your Confluence connection.

      • Confluence Personal Access Token: Your personal access token for Confluence.

      • Wiki URL in Confluence: The base URL of your Confluence wiki (e.g., https://yourcompany.atlassian.net).

    • For a Web connector:

      • Connector name: A name for your web connection.

      • Base URL: The starting URL of the website you want to crawl.

      • Web connector type:

        • Single: Only crawls the specified base URL.

        • Recursive: Crawls the base URL and all linked child pages within the domain.

      • Scroll before scraping: This option ensures the agent scrolls through the entire page before extracting data.

  5. Click Submit. Once added, AI Connectors automatically refresh their data every 30-40 minutes (default frequency) to ensure the agent has access to the most up-to-date information.

Using the AI search agent​

Once your AI Connectors are set up and actively indexing data, you can start using the AI Search agent.

  1. Navigate to the Agents page and click New chat (defaulting to the AI Search agent) or navigate directly to the AI Search agent from the agents page.

  2. Apply filters (optional): If you want to limit your search to specific sources, click on the Sources filter and select the desired AI Connector types (e.g., "Web" for searching only web connectors, "Jira" for Jira data).

  3. Type your query: Enter your question in the chat box.

  4. Review the response: The agent will provide an answer based on the indexed data. It will also list the source documents that contributed to the answer, indicating their relevance (e.g., "D1: Customer Service is Solid Health Customer Support and Service - 88% match"). You can click on these source references to view the original document.

Examples of usage:

  • "Show me recent tickets in Jira." (If a Jira connector is configured).

  • "What are the key findings in the Q4 sales report?" (If a "Q4 sales report" file connector is uploaded).