› Workflow › Business processes › Design a basic process › Add an activity › Activity types › AI Knowledge Base activities › Add to knowledge base activity › Configure an Add to knowledge base activity

Configure an Add to knowledge base activity

Use the following procedure to configure an Add to knowledge base activity.

The Add to knowledge base activity is only available to Enterprise-tier customers. For users on the Standard and Advanced tiers, this feature is restricted and indicated by a Lock icon.

  1. Add an activity to the map. See Add an activity.

    By default, the Configuration tab is open in the properties panel of the activity.

  2. Enter a Name for the activity.
  3. On the Type list, select AI Knowledge Base > Add to knowledge base.
  4. By default, the Knowledge Base Thread Pool is selected in the Thread pool box.

    When you import a process from a version lower than TotalAgility 2026.1, the Thread pools field is blank by default. You can add a thread pool.

  5. On the AI Knowledge Base provider list, select the knowledge base provider configured. See Integrate TotalAgility with AI Knowledge Base.

    If using the Tungsten knowledge base provider, you must have the required licenses to add documents to the knowledge base.

  6. The indexes that are added while configuring the AI knowledge base integration are automatically added in Azure AI Search and appear on the Index list. Select an index.
  7. On the Input list, select one of the following process or server variable types.

    Variable type Description
    Document A single document is added to the knowledge base.
    Folder All documents within the folder are added to the knowledge base.
    String/Data object

    The text provided is added to the knowledge base. You can access the unique ID for any non-document data added to the knowledge base and use it to delete that data.

    The Mime type, Sources, and “Overwrite if exists” properties are not available for String and Data object input types.

  8. On the Tag list, select a process or server variable of type String or provide an inline value such as SharePoint or an external system.

    The Tag property is only available if you select the input variable as a String or Data object.

  9. On the Mime type list, select a mime type, such as PDF.
  10. Under Index fields, you can index fields to a document and modify or delete existing index fields. You can add multiple fields; however, you cannot use the same index field type more than once.
    1. Click .
    2. In the Name box, click the down arrow to expand the list, and select an index field.

      The Name list includes the index fields defined in the AI Knowledge Base integration in addition to the Job ID,Case ID, and Case Ref system index fields.

    3. In the Value box, click the down arrow and select a server or process variable, Agent (synchronous process), or custom service as the value of an index field. The synchronous process should contain a single input (String) and a single output of the correct type for the field.

      If an index field is vectorized, you can only select string variables.

    4. Click Add.

      To edit a filter field, select the field and click and modify the required fields. To delete a field, select the field and click .

      When you upgrade TotalAgility or import a process from a previous version, all filter fields are automatically set to "Equals", and the overall filter operator is set to "And".

  11. To overwrite a document or all documents in a folder if they already exist, select Overwrite if exists. (Default: Clear)
  12. To suspend the job if the document already exists in the knowledge base, select Suspend job on error. (Default: Clear)

    The "Suspend job on error" property is only available if the "Overwrite if exists" property is not selected.

  13. Save the process.

    When the activity is executed:

    • The documents and text are added to the knowledge base with the configured identifiers.

    • If you select an Agent (a synchronous process) or custom service, a job is created for each chunk on the synchronous process. The chunk text is processed to generate an output value, which is then used to populate the index field.

    • The document/data is added to the knowledge base, and the mapped variable (String or Data object) is populated with the unique ID assigned to the chunk in the knowledge base.

    • An audit log entry is recorded.

    • An error occurs if the document does not exist and the "Overwrite if exists" property is enabled.

    When uploading large documents to the AI knowledge base, you may need to increase the target activity duration according to the document size to ensure successful uploads.