Daniel Dines, Chief executive officer and founder at UiPath 
Daniel Dines, Chief executive officer and founder at UiPath 

RPA is about taking robot out of humans: Daniel Dines

  • In RPA, software robots mimic and integrate human actions within digital systems to optimize business processes 
  • A process would need lesser people to handle when RPA is applied. But it does not automatically render workers jobless

NEW DELHI: Robotic Process Automation (RPA), is a buzzword in the enterprise market today. Research firm Gartner expects the RPA software market to touch $2.5 billion by 2022. In an interview, Daniel Dines, CEO and founder of global RPA vendor UiPath, explains why RPA is taking the enterprise world by storm. He also touches on how RPA is being combined with artificial intelligence (AI), machine learning, natural language processing (NLP), and many other emerging technologies to deliver more value to companies. Edited excerpts:

What exactly is RPA? And how does it benefit companies?

In RPA, software robots mimic and integrate human actions within digital systems to optimize business processes. RPA automation captures data, runs applications, triggers responses and communicates with other systems to perform a variety of tasks. These robots interact with applications and systems through a graphical user interface or command-line interface to carry out routine tasks.

Can you give us some examples?

A new credit or loan application, for example, requires data to be accessed from a hand-written form, bank documents in PDF format, credit score which is accessed online, etc. If these steps can be automated through RPA, you can process credit applications faster, leading to higher efficiency and lower costs.

How does the global RPA market look?

RPA has been one of the fastest-growing markets in enterprise technology. And analyst organizations have been revising RPA growth numbers upwards. Everest Group pegs the market growth rate at 75-90% between 2017 and 2019. Forrester Research raised its market guidance for RPA to a $3.3-billion market by 2021. UiPath has been growing faster than the market.

What impact will RPA have on jobs done by humans?

RPA helps humans by taking out the drudgery of monotonous tasks and frees the human to do higher value tasks. At a unit level, a process would need lesser number of people to handle when RPA is applied. But it does not lead to those workers being automatically rendered jobless. It is like understanding the difference between driving a car with manual transmission, driving an automatic car, and driving a self-driving car—the driver is there in all cases.

Not all processes can be automated to an extent that it does not require any human intervention. In the RPA world, we call it “attended RPA", which includes scenarios where decision making and/or user input is required, such as desktop automation. These software robots work at an employee’s workstation and are triggered by either a user’s command, or the robots need input from the user to continue a task.

Are you seeing RPA merging with AI to deliver more business value?

Yes. RPA is closely tied to AI. In many ways some of the AI technologies have been at the core of an RPA platform like UiPath. Our strategy is to continue to invest heavily to deliver automation embedded with AI technologies.

Would it be right to call voice- and chat-enabled bots the next wave of RPA?

Chatbots are included in the RPA platform. So, I wouldn’t call it the next wave. Chatbots are part of an area called conversational AI. Chatbots, virtual assistants and bots are closely tied in. While a bot only follows the script, the chatbot and virtual assistant have more options. Supported by AI, they understand the meaning of what was said or typed.

Critics say RPA products need to sharpen focus on unstructured data. Your comments?

Right. Data is the starting point. And that’s what UiPath’s focus has been right from the beginning. Unstructured data in the form of images/documents is captured through OCR (optical character recognition). Unstructured text requires NLP to understand context, entities, person, place, etc. Unstructured images and video require vision technologies to convert them into data. Unstructured audio requires speech recognition.