Run AutoML Business Process

After you activate a Cognitive Bot by configuring AutoML in a Manual Task and run the Manual Task, Control Tower automatically launches an instance of the out-of-the-box AutoML Business Process. This BP is auto-created, with the original Manual Task inside it used for those cases when manual processing is still required.

Now let's work on the AutoML Business Process configuration.

Configure AutoML BP

In the Business Processes list, select your newly created AutoML BP and perform the following steps:

1. On the Data tab, click Upload data and map the input column(-s) with answer types, if needed.

2. Configure advanced AutoML BP settings.

3. To set the model execution options, navigate to the Workflow tab and double-click the Manual Task in the flow.

   

4. Navigate to Configure AutoML and select Execute Model.

   

note

By default, the Machine Learning model, Cognitive Bot Workflow, ID, and Version parameters are set according to the training configuration of the initial Manual Task.

• ID is the unique identifier for the experiment group of a trained model (set during model training) in the [answer unique code]_runUUID format. You can select another trained model ID from the list, and Cognitive Bot applies the selected trained model for execution. This list is not sorted by the Machine Learning model type.
• Version: it is the version of the selected Machine Learning model. The list consists of available Machine Learning model versions. The default value is the version used for the selected Machine Learning model during training.

4. Run your AutoML BP.

caution

Input column(-s) mapping may not work in the AutoML Business Process after activating the Cognitive Bot in the original Manual Task. To avoid this, provide input data files with column names that match answer codes.

Optionally, you can edit the AutoML BP bot steps, for instance, enable or disable Auto Quality Control (AutoQC), set data formats, and so on.

Reuse AutoML BP

If you have additional input data, you can reuse the AutoML BP in one of the ways described below:

• Go to Manual Task, set new accuracy parameters, and click Apply Recommendation.
• Copy and run the new AutoML BP.

Monitor AutoML BP

You can monitor your AutoML BP progress in the BP workflow diagram in the view mode or on the Data tab.

In the sample workflow diagram below, 24 documents were uploaded as the input data with the following results:

• 18 documents were successfully extracted by the Cognitive Bot and normalized.
• 6 documents were routed to human workers because the Cognitive Bot failed to extract information.

note

You still need to complete some number of Manual Tasks in the following cases:

• Automatic extraction failed or was only partially successful.
• Data normalization failed.

Finally, you get the resulting data partially generated by AutoML, partially submitted by human workers.

What's next?

• To improve the results, apply post-processing.

• To enable analytics, follow the instruction.