When process automation is called in IT applications, traditional examples add filters to our email inbox or salespeople who configure complex email follow-up campaigns when inbound emails come in. However, companies like UiPath are moving process automation to the next level. They offer a suite of solutions that empower developers and non-developers to construct practical software robots that can measurably improve business efficiency. In this interview, we speak with Boris Krumrey, Global VP Automation Innovations at UiPath. We overview the platform and discuss the current and future state of Robotic Process Automation(RPA).

Business processes are not straightforward to optimize. Like hiring a new employee, there are many steps and different systems involved. Imagine tethering these all together with the UiPath automation platform and turning them into an application in a single workflow. 

Boris Krumrey says, "in one application, one UI, where you enter all the data, and it will then do the entire workflow, sending the email, getting the approval, understanding natural language, even extracting, if necessary..."

Traditionally, the hiring process involves filling out many forms and giving them to HR, getting approval, and opening new accounts. So, for example, when HR needs to provide a new laptop to the person, a new engineer needs to fill out many forms, and he needs to put that data into the backend system of HR. Then, he went into the internal IT purchasing system to order a laptop. So there are 5-6 different interfaces, from APIs to UI. 

UiPath converts this to a straightforward process that can be used as a toolkit. You can build an API on anything, even if the mainframe application does not have an API at all. It is flexible to connect to anything. Recently, Uipath also extended its platform for full API management capability. They acquired Cloud Elements company, and they received features like ready-made API. You can drag and drop it into your workflow. There is a lot of room for multitasking as well. You can define triggers and say, Okay, if this SAP system table here is being updated, then I want you to update another table in another system. It is a new way to approach and connect those systems. If there is an API and only a mainframe, you have to go to a different user interface and combine API with UI automation in any workflow. 

Uipath works differently than many of the platforms. For example, let's say there is software in a company written by someone who has not been in the company for 15 years; it is in a language nobody knows. In such problematic situations, UiPath could automate the system as well. When company workers use that software, Uipath staff generate the automated workflow. They record data input and output in that system so Uipath can integrate that process into its workflow. In this way, you can modify, adjust and make it flexible, put parameters, variables, and whatever around it as you want, and turn it basically into a complete application. Later this application runs onto a data entry person's machine, or you also can have a separate virtual machine with a robot configured that works side by side. With robot implementation, it would perform in an environment without any attendance. 

Uipath also has automation for more hard to automate things like natural language processing or computer vision. They have embedded computer vision features and a whole AI engine, which means they can take any machine learning model that you've written in Python code. You can upload it into the Uipath engine and execute it within the workflow. That's immensely powerful. You can use it for fraud detection and any prediction you want. 

It provides the whole set of ready-made solutions, whether it's particular models for predictions of defaulting loans or their implemented solutions for customers. For instance, they did the prototype of the cash collection process. These things predict how many customers and which customers will not probably pay in time, their bills, and so forth. It can also trigger specific actions and works as a reminder. So briefly, it can help to control a company's cash flow. The ready-made models they have prepared also work on this concept. The robots can extract information from any forms, invoices, and receipts sent to them. They use these processes as what they call humans in the loop. Whatever the system can read, if there is anything that the system is not confident to recognize, it just then puts it into the queue of the human to revise it. 

So when the system revises it or makes any correction, it would track that information and then retrain the model automatically, so automated ML is also part of the platform. Boris Krumrey says, If you think of a small business or a big business, they all have to use certain same key functions, yes, to a different scale. There's always the swivel-chair type of activities everywhere. Whether I have to process thousands of invoices, or whether I have to process in a small company and there are hundreds of invoices every month, it's still a lot of work. If I can automate it and save that time, it is as impactful for small businesses and as for large businesses. 

RPA is not all about cost savings; its other main advantage is increasing productivity. It gives you more capacity to do work that humans are doing and not the work that robots should be doing, which is valid for whatever you are doing. Since we are humans, we could forget some data or mistype something, and correcting that would take even more time. All those things can be eliminated by using RPA. So people focus on things they are best at, like creative workloads. Thanks to cost savings and increased productivity, it creates a win-win situation for both employers and employees. 

It would be ideal to have a background in software engineering to take full advantage of the platform. Because UiPath has now grown to the capability of being a whole application development platform. You can combine all these automation flows into an actual application that you can trigger within. You have a user interface on top of all this automation. When you do more extensive projects, advantages become much more apparent. 

On the other hand, they also target non-technical users. It could be users operating a process, for instance, an accountant, a sales operator or a customer service agent. If you are interested in building something logical, you can create whole automation yourself. Uipath has a special product called Studio X. Studio X is designed for such typical user needs. A typical user needs to extract data from place A, put that into system B and then extract another data from system C and combine that. So ideally, for all those things, anything you do around Microsoft productivity tools, or Google productivity tools, you can combine and turn into fully-fledged automation. 

Sometimes, the laptop you regularly order could be out of stock, and monitoring all these could be a tedious task. The exception handling feature is also one of the things Uipath makes very successful. It has a process mining tool that links all these systems and gives you a log of what people are doing. What helps with visualizing an entire flow is showing you how many deviations of this end-to-end process exist. With that information, you could identify whether a process is highly standardized or not. If it's not, you know that your automation journey will take much longer. There are tools available in the platform that help you discover and assess how you're doing certain things. 

You have an excellent interface that would alert you if anything goes wrong. If you come across an exception, you could pick that up fairly quickly. You can use UiPath to test and automate the tests for any kind of application; web applications, mobile applications, whatever your software development target is. For UiPath, automation is intelligent; there is no difference between intelligent automation and just automation because it's such an integral fabric and an integral part of applying decision models into the entire workflow. 

This article is based on an interview with Boris Krumrey, Global VP Automation Innovations at UiPath. To listen to the full interview, click here.