Machine Learning Workbench

Automation Studio
1

Hi Team,

Can you explain how the Machine Learning Workbench is used in Contextual Intelligence in AssistEdge RPA?

2

With AssistEdge Enterprise 19.1 - Decision Workbench, users can create a Automation Process that requires human validations - like approvals, escalations and exceptions.

AssistEdge Decision Workbench has capability to learn from human decisions, It comes up with Machine Learning model that is used to build the business decisions, that can be marked to be reviewed by Process - Subject Matter Experts before being used for automation.

You can refer knowledge base for Contextual Intelligence and Decision Workbench -

https://www.edgeverve.com/assistedge/knowledge-base/RPA19.1/Automation_Studio/Configuring-Automation-Processes/Contextual_Intelligence/Contextual_Intelligence.htm?rhhlterm=machine%20learn%20learnings%20learning&rhsearch=Machine%20Learning

3

AssistEdge Enterprise 19.1 - Contextual Intelligence is applied to automation process to have ability to adapt and apply what is being learnt in multiple scenarios. To understand How Contextual Intelligence works see below scenario.

Scenario - User need to implement a scenario in automation process, where AssistEdge RPA robots predicts when industrial drilling machine requires maintenance or not, based on available historical data.

For more details on configuring contextual intelligence in your own process, You can refer knowledge base with step by step guide -

https://www.edgeverve.com/assistedge/knowledge-base/RPA19.1/Automation_Studio/Configuring-Automation-Processes/Contextual_Intelligence/Step_by_Step_Guide_Decision_Model_and_Decision_Workbench.htm

4

Hi,

Additionally AE RPA provides integration with H20 an open source, machine learning platform. where in you can call built in ML models and can be used for prediction and other ML specific use cases

Regards,
Amit

5

Adding on to @amit_anand15 and @sumit.sagar response

@mrunalini

It is not mandatory that you should use the workbench only for ML related use cases. You can use this Workbench whenever you need to bring human into the loop.

For example -

If you’re doing PDF extraction and want to ensure that no erroneous data is getting extracted, you can validate that by adding Decision Workbench activity.

You can make the human verify if the data is correct, modify the values if needed, and submit it.

Upon submission, the subsequent process will be called and right data can be added into the downstream application.