How Artificial Intelligence Powers Celonis’s Action Engine

When you search for “artificial intelligence” in most stock photo websites, you get a bunch of photos of robots working at desks. Because most people don’t understand how to work efficiently with AI-powered systems, there’s a pervasive fear of actual robots stealing our jobs, as if companies undergoing digital transformation were simply taking laptops out of employees’ hands and giving them to Rosie from The Jetsons. Not that anyone at Celonis would hate having a robot coworker; we’re just not there yet.

The closest thing we’ve got to a robot assistant is our Action Engine, a sophisticated business optimization program that uses machine learning to solve problems for companies and increase efficiency. No, it’s not a wisecracking AI like Futurama’s Bender, but it also doesn’t have a gambling problem, so you win some, you lose some.

What is the Action Engine?

Our Action Engine is the central hub of innovation at Celonis; it’s the intelligent system that processes all the customer data we collect during the discovery phase of Process Mining, and it makes continual suggestions for places companies can improve. In short, it’s the UI that enables companies to work differently. Marcell Vollmer, Chief Innovation Officer at Celonis, says one of the goals of the Action Engine is to “set the standard on UX for users, so that they actually want to use the software. Our customers tell us they’re using countless programs every day. We don’t want to create yet another tool for them to use. It has to provide additional value.”

So, how do you solve the “too many programs” problem with yet another program? Well, you train the Action Engine to simplify every process, whenever possible. That means the intelligent system keeps the priorities of every stakeholder in mind, from executives down to employees who put in the most hours on the program itself. 

You may have heard it said that the best UI is no UI, meaning that humans should ideally never have to interact with or recalibrate the technology they use to do their jobs. We believe something slightly different. At Celonis, artificial intelligence augments the human workforce. The Action Engine doesn’t just eliminate the need for human workers to complete rote or boring tasks; it identifies opportunities for those workers to innovate. The goal isn’t a reduction of headcount across a company; it’s ensuring that every employee’s gifts are being used most effectively, by pairing each complex task with exactly the right mind to do it in ways an AI can’t yet imagine.

How artificial intelligence speeds up process improvement

Using artificial intelligence, the Action Engine offers something unprecedented to modern businesses: analytical insights put to work in your day-to-day operations. Our AI monitors every open case (i.e. order, invoice, ticket) and can predict it likely outcomes, recommend steps to help reach your goals, then help you execute those steps automatically. It’s like having Google Maps for your business processes — it proactively finds the traffic jams and road work then suggests an alternative route to get you to your destination. Because Celonis strives to perfect the art of Process Mining, the Action Engine spends most of its time reducing friction and frustration in the workplace. It powers our investigations through each of the first three phases of Process Mining: discover, enhance, and monitor. Because the Action Engine never ceases its search for the best business process, it’s always improving. That means your business will keep getting better as time goes on.

Process Mining arms professionals with artificial intelligence

Essentially, Celonis allows us to automate the basic, transactional tasks that humans don’t enjoy doing (for good reason!). However, it also requires a team to oversee the implementation of Process Mining across an organization, arming them with the tools they need to improve their workplace. In that sense, we believe that automation will end up creating more jobs than it eliminates, and those positions will require more human ingenuity than any of the rote positions it turns into dust.

Thanks to systems like the Action Engine, we foresee a world where more people feel genuinely curious about what they do at work. Because prioritizing business outcomes, making associations between patterns of behavior, and motivating teams are all rewarding forms of labor that AI won’t take away. After many years of development, the process improvement technique that professionals have been waiting for is finally here.

Emily Gaudette

Writer