Blogs • 24 Nov 2020

A deeper dive into Intelligent RPA

This blog post aims to follow on the last one about IRPA. The previous post gave us a global view on what IRPA is generally about, this post aims to dive more into how it works, diving into the more technical aspects of the solution. It is going to give you insights into the power and possibilities of SAP Intelligent Robotic Process automation.

What is SAP iRPA and why to choose this solution over its competitors?

When SAP acquired Contextor it got renamed to intelligent Robotic Process Automation. SAP added some features that make the solution interact much further than it did in the past. The reason why the word intelligent was added is simply because it is extending the capabilities of what the bot can do. By connecting the bot to APIs, you can make the interaction very close to what a human can do. Also, by using AI and machine learning the tasks that the bot is doing are improved, it’s continuously learning and analysing.

RPA has many benefits, but how can one solution be better than the others? Well to elaborate on how and why SAP’s IRPA is different from its competitors, let’s compare it through these 6 different categories: Functionality, Compatibility, Comprehensiveness, Prebuilt bots, Security, Pricing.


  • There are 2 types of bots: a digital worker and a digital assistant. The digital worker is an unattended bot meaning that it works like an employee does, doing its job without the need of continuously checking what it is doing. On the other hand, the digital assistant is working with you. It helps reduce the number of steps it would take you to complete your job.
  • Everything is done in a single solution: you can design, configure, orchestrate, run, monitor and debug your project inside a compact solution.
  • With the help of a drag and drop designer, you can build a complete bot very quickly. Choosing from a series of pre-built functionalities you can find what you are looking for with ease.
  • In the situation where you might want to configure functionality that does not yet exist in a prebuilt package, there is also a script editor that lets you design custom functions.


  • Obviously, because it is an SAP solution, it offers an unrivaled ability to drive SAP applications.
  • It also has the ability to interact very well with non-SAP solutions like Microsoft Office applications (Word, Excel, Outlook,...).


  • Building an IRPA project is done through a series of tools. The Cloud Factory offers a very detailed dashboard to monitor what your bots are getting up to: are they doing what they are supposed to be doing?

Prebuilt bots:

  • There is a library in which you can find a large amount of prebuilt bots to speed up the development of bots built for e.g. SAP S/4HANA , SAP SuccessFactors,.. They are built following SAP’s best practices.


  • User credentials are stored in a dedicated repository; So only people with the right clearance have access.


  • Flexible pricing. It follows the principle of you only pay for what you use. 

We can now see that iRPA can do many things, most of these functionalities are only found in SAP’s solution. This is what makes it a very powerful tool.

How SAP IRPA works

Using SAP Intelligent Robotic Process Automation (SAP iRPA) can help in automating processes partially or fully. This would allow employees to focus on tasks and activities with higher added value. All the means to reduce the time employees must spend on tedious, error-prone, low-value tasks.

SAP iRPA consists of 3 different components:

  • Desktop Studio
  • Desktop Agent
  • Cloud Factory

They work together in giving a streamlined experience in both the development and monitoring of the working robots.

Desktop Studio

The Desktop Studio is where the development process happens. This tool allows us to teach the bot the different steps that it needs to complete each scenario. It is done in 3 steps:

  1. Capture the different screens that the bot needs to navigate through
  2. Determine the order of the steps
  3. Clarify what it needs to do at each step by writing the necessary code

When the bot has been “taught”,  it’s onto the testing part. With the help of a powerful debugger, all the testing can be done through the Desktop Studio. This component is solely used for development. When the bot is fully working, the next two components are used to manage the bot in production mode. 

Desktop Agent

When the development is finished, the Desktop Agent is downloaded on your system and works inside it, continually executing the configured steps. Depending on the configuration of the scenario,  the Desktop Agent is either launched by a user or it’s triggered by a process. A digital assistant and a digital worker are the two ways the Desktop Agent can work. The digital assistant works alongside a human, helping to increase the speed of completing the task by prefilling fields. The digital worker does a complete task on its own without the need for human intervention.

Cloud Factory

The Cloud Factory is the heart of the management, configuration and deployment of the different automation packages. Because it is cloud-based, all configurations are securely stored independently from your on-premise systems.  All user credentials are stored in variables in the Cloud Factory, meaning that only the users that have access to this component can see this information.

The included monitoring tools allow you to see what your bots are doing.

The Cloud Factory provides access to a store full of pre-built packages which you can use to get a bot up and running very quickly.


Summary rpa

When development is finished in the Desktop Studio, you export the steps in a package. These packages are deployed to the Cloud Factory. The packages are then provided by the Cloud Factory to the Desktop Agent.

Automate SAP with IRPA

After seeing all this content it may be a little overwhelming and blurry. To delve into a deeper understanding of the subject we are going to walk you through building a real-life scenario and explaining how these components work. 

Before jumping into how we built the demo, we would recommend seeing it in action first. We have created a small video to do so.


As seen in the video, here is an overview of what it does:

sales order creation

The bot has 4 steps to complete. Firstly it is going to read an outlook mailbox and see if there are any new orders that are coming in. If there are, there will be an excel file attached. The bot then downloads the files and reads the values from inside it. It then needs to use these values to create a sales order in SAP S/4HANA, when everything has been completed successfully it sends a confirmation email.

As we saw in the previous chapter, the first step is developing the bot. We start in the desktop studio defining the different screens and input fields. This step really consists of teaching the bot what it has to do.  The steps tend to repeat themselves after a short amount of time, so to keep it short and concise I will focus on creating the sales order in S/4HANA

In the image below, one of the steps is to create a sales order in SAP S/4HANA. We are defining the fields which the bot needs to fill in.


Then we have to determine what the bot has to do with all these screens that have been captured, which is done following the logic of a flowchart.


Now each flow of the chart needs to be configured. This is done by code, which represents the last step before seeing if what you have built works.

With all projects in their development phase, there could occur some errors. Most of the time these errors are rather hard to understand exactly where they are coming from, this is where the help of a powerful debugger comes in handy. The desktop studio has you covered for this step.


When you start testing, the screen that you can see in the image above pops up. It allows you to do several things: log each step that the bot has completed, show if the elements on the page are recognised and it allows you to test snippets of code directly. This last feature is very useful, you can run new code directly from inside the debugger without needing to restart the scenario from the beginning. With the combinations of all these features, it puts you in a good place to test your bot correctly.

When the development has finished, it’s now time to go into the production phase. You can’t see what the Desktop Agent is doing because he is working in the background. The way to visualize its activity is through the Cloud Factory. It offers a detailed dashboard to show when it’s online, how long it’s been working and some statistics.


To help you discover the power of SAP Intelligent RPA, we have developed a number of "Challenge Accepted" proof of concept packages. These give you the opportunity to quickly implement and test your own use cases. Of course, we can also develop software bots completely tailored to your business needs.

Ready for your first RPA success story?

Speak to our team and learn how you can benefit from RPA.


Blog by Nicolas-Tate Smith, SAP Technical Consultant