Skip to main content
One of the key strengths of agents is their ability to improve over time. Cognigy offers the intent trainer feature that allows you to refine agents based on existing dialogs. This tool improves the agent’s understanding by letting you review user inputs and add them to Intents.

​
Key features

  • Analysis of collected user inputs. Review and analyze intent trainer records to identify areas for improvement.
  • Adding inputs to intents. Choose which intent trainer records enhance specific intents and add them to the example sentences of those intents.
  • Instant training. After refining intent trainer records, quickly update the NLU model on the Intent Trainer page. This action eliminates the need to click Build Model on the NLU page.
  • Automatic scoring. The Train button automatically scores intents, and you can visually track progress with color-coded icons and scoring data in the list.

​
Limitations

  • By default, the time-to-live (TTL) of intent trainer records is 10 days (14400 minutes). If you have on-premises installations, you can change this value using the following variable: TRAINERRECORD_TTL_IN_MINUTES. For example, to set the TTL to 30 days, configure the variable as follows: TRAINERRECORD_TTL_IN_MINUTES="43200".
  • The maximum file size for uploading intent trainer records is 150 MB.

​
Work with the intent trainer

In Advanced > Intent Trainer, you can:
  1. Filter intent trainer records
  2. Manage intent trainer records: add to intents, skip, or ignore
  3. Train the NLU model
  4. (Optional) Import and export intent trainer records

​
Filter intent trainer records

Review the collected input records from users and search for them using filters. To search for a specific user input, enter text in the Search utterance field. The list shows the Utterance, Count, Intent and score, Last edited, and Review status columns. If the NLU didn’t find an intent for a record, the Intent and score column shows No intent found.

​
Manage intent trainer records

Based on the analysis, decide which user inputs improve a particular intent, and add those inputs to the corresponding intent. The Review status column in the list is a drop-down for each record. When you select a status, the change is marked as unsaved. To apply all unsaved changes in one step, click Save in the upper-right corner of the page. The button shows the number of unsaved changes, for example, Save (3). To discard all unsaved changes, click Cancel.
  1. In the Intent Trainer list, in the Review status column of a record, select Assigned to intent.
  2. In the Add to intent dialog, select a Flow and an Intent to which you want to add the record.
    If needed, change the text of the record in the Utterance text field.
  3. Click Add to intent.
  4. Click Save. Cognigy adds the record as a sentence to the selected intent.
To create a keyphrase or a synonym for a record, click the record or select Edit in the menu of the record to open the utterance page. In the Lexicon section, enter a keyphrase in New Keyphrase and select a Slot, or enter a synonym in New Synonym and select a Keyphrase. Then click Save. Cognigy creates keyphrases and synonyms immediately, while the review status of the record stays unsaved until you click Save in the list.
  1. In the Intent Trainer list, in the Review status column of a record, select Skipped.
  2. Click Save. This action moves the input to Skipped records, but it reappears in Not reviewed if the same input is entered again.
  1. In the Intent Trainer list, in the Review status column of a record, select Ignored.
  2. Click Save. This action moves the user input to the Ignored records, and if a user enters the same input, it’s also ignored.

​
Train the NLU model

After you save records that you added to an intent, click Train next to the Search utterance field. The Train button is available only after you save records that you added to an intent. You don’t need to click Build Model on the NLU page. The intent trainer trains the flows that received new sentences and shows the result in the scoring data and color-coded icons.

​
Import and export intent trainer records

You can import and export records between production and development environments.
The intent trainer temporarily stores records. To ensure data safety, export the records for local storage.
  1. Go to Advanced > Intent Trainer.
  2. Next to the Search utterance field, click upload (Import Trainer Records).
  3. Select a file in the CTRAIN format from your computer and click Open.
Once the file is uploaded, you’ll see the Trainer Records imported. message.
  1. Go to Advanced > Intent Trainer.
  2. Next to the Search utterance field, click download (Export Trainer Records).
  3. In the Export Intent Trainer Records dialog, select a date range under Select Date range by clicking the date and selecting the desired date in the calendar. By default, the range covers the last seven days.
  4. To include reviewed records in the file, activate Include reviewed.
  5. Click Confirm, then Download Trainer Records.
The file is downloaded in the CTRAIN format.

​
More information

Last modified on November 11, 2026