In this session, we take a look at defining intelligent automation and introduce a new (free!!) tool to help with creating and consuming a Machine Learning model.

Summary

1. Robotic Process Automation (RPA) is task centric, rule based automations that take place across application programming interfaces (APIs) and graphical user interfaces (GUIs)
   1. Artificial Intelligence is very broadly defined as data-driven modeling mimicking cognitive decision making.
   2. The sweet spot where these two technologies can overlap - where RPA bots consume machine learning models - is intelligent automation
2. There are several examples of intelligent automation that many RPA developers are familiar with (and possibly already using)
   1. IQ Bot - intelligent document extraction using IQ bot makes using of machine learning to identify, cleanup, and extract text from invoices and other structured or semi-structured forms
   2. Discovery Bot - process discovery, process recording, and recording comparison all enable Discovery Bot to intelligently work to build bots automatically as the output of submitted process recordings
   3. Chat Bot Integration - many organizations use a chat bot front end to gather information from internal or external customers, and use that information to trigger and send data to bots for automated task processing
   4. Integration with 3rd Party Services - available cognitive services like those available from Microsoft Azure, Google Cloud, and Amazon (among others) allow bots to consume cognitive services directly from within a bot. Several bots/packages supporting such integration are already available on Bot Store.
3. The Machine Learning Classifier utility introduced allows anyone (developer or otherwise) to build, test, and consume a machine learning model using pre-classified data that they have on hand.
   1. In the video, we used a popular ham/spam data set to train and test a model for its accuracy. In part II of our series, we'll be giving everyone access to the application so they can build out a machine learning model of their own.

Learn More

Take a look at Intelligent Automation Pt II: Building the Bot if you're ready to move on. In this session, you'll get hands on with building your own matching learning model using the tool Micah and Harish demonstrated here, and the session also takes users through building a full bot to consume the model. You won't want to miss it!