Frequently asked questions – Decision

Blue Prism® Decision provides Blue Prism users with the ability to train and deploy intelligent machine‑learning decision‑making models within their Digital Workforce, without requiring any data science expertise. Blue Prism Decision's simple and intuitive user interface allows users to:

• Define a model – Add a title, description, decision type, and decision variables.
• Train the model – Use active learning based on defined input variables specified by the model service, without needing any data science knowledge to produce statistically accurate machine‑learning models.
• Calibrate the model – Review model inputs and outcome predictions, and overwrite model decisions if required.

Decision is designed for Blue Prism process developers, to build machine learning models without any data science or statistical knowledge required.

Decision is a no code AutoML capability – it has been designed to be easy-to-use without the user needing to have specific coding or statistical knowledge. However, experience in Blue Prism process development is required for integrating into a Blue Prism process and a thorough understanding of the inputs and outputs of the use case you are automating is needed to train an appropriate model.

Blue Prism® Decision is currently available to customers who have Production or Business Critical support. Contact your Blue Prism sales representative for details.

Yes, Blue Prism® Decision does require a license.

Decision can be downloaded from the Blue Prism Portal.

As a simple overview to the Blue Prism® Decision installation process, you will need to:

1. Create the SSL Certificates required for the install.
2. Install the Blue Prism Decision container.
3. Install Blue Prism Hub.
4. Log in to Hub, install the Decision plugin and set up the user access.

For the installation details and prerequisites, see Installing Blue Prism Decision and Installing Hub.

Blue Prism recommends that Decision is only installed if you have a mature understanding of IT policies around Docker Container infrastructure, as this release is dependent on installing both Hub and a Docker Container. The next release in 2022 will also be made available via an MSI.

Yes, the Blue Prism Decision container requires a Docker host. For more information, see Installing Blue Prism Decision.

Yes, however, you will need to update the settings in the Hub appsettings.json file to do this. Changes to the appsettings.json file should be made in conjunction with Blue Prism Support to ensure your system is supported. For more information, see Troubleshooting the installation.

Yes, the models trained and calibrated in Decision can be integrated into your processes in Blue Prism via the Decision VBO. For more information, see the Decision user guide.

Decision supports Blue Prism 6.4.0 or later.

The version of Decision that you are using is visible in the footer of the web page. Hub administrators can also view the version number on the Plugin management page in Hub. The plugin versions currently all use the same version number as Hub.

A notification displays to users in Hub when their license is within 30 days of expiry.

The following languages are supported by the Decision user interface:

• English (United States)
• French (Français)
• German (Deutsch)
• Japanese (日本語)
• Spanish (Latin American) (Español)
• Simplified Chinese (中文)

For more information about localization, see Localization.

Decision uses active learning and AutoML. For more information, see Decision model overview.

To use Decision to the best advantage:

• The data must be tabular (in a Blue Prism collection).
• It is recommended that the use cases contain between 3-8 input variables, however, there is no limit to the number of variables you can use.
• The current decision-making process in the use case should be thoroughly understood. The better the understanding of the process and the considerations that are used to reach a decision, the easier you will find to model it in Decision using variables and training.

Decision stores any inputs and outputs that are used when a model is used by Blue Prism. The prediction list displays the unique Correlation ID associated with the automation that used the model, the inputs provided, the model outcome, and the date and time this occurred. For more information, see Viewing a model's usage.

Yes. Please refer to the product roadmap on the Blue Prism Portal for more information on plans for future releases.

Blue Prism provides an ideation portal which is accessible through the Blue Prism Portal. New ideas as well as the rating of existing ideas can be accessed from here.