Increasing trust in robots and unburdening colleagues

Rianne Selhorst, Team Lead & Product Owner Robotics & AI at Rabobank

"Three years ago, nobody really knew anything about robotics. So we figured it out for ourselves. Within Finance, our robots have now done the equivalent of 45 full-time jobs. My job? Helping colleagues to develop a new mindset: let robots do the tedious work, so that you have more time for the real and fun thinking'.


Robots love boring

Colleague Harry is one of our most diligent colleagues, he even picks up important customer data at night. As a robot, he takes over simple, often repetitive tasks. Among other things, our robots can handle financial processes, register invoices and prepare reports. They do this 24 hours a day from their own virtual workplace, because they are not physical robots but they click through systems, so they can search, collect and organize. Robots help us work more efficiently and they take over the most boring jobs. A good example is the robot that does all the boring research when a customer wants to close an account. Colleagues can then put their energy into customer contact’.

Forerunner in Robotics

With Robotics, also known as Robotic Process Automation (RPA), our team is now a forerunner in the Netherlands. That’s remarkable, considering that three years ago none of us knew what it was because it was very new at the time. We could hardly find any outside experience, nobody had worked with RPA for more than a few months. So we had to find out for ourselves, driven by our desire to learn every day and do things better. We experienced that you can learn RPA very well. As long as you have an affinity with IT, can think analytically and have a constant need to improve. We started at Finance with three people on RPA, including me. And it actually came about organically, because we were already setting up our Finance processes more efficiently. RPA was the logical next step.

You have control over robots, they do exactly what you make.
More on Digital & Innovation

Secure virtual colleagues

Our first robots have now retired. Like the robot that recovered wrong customer transfers for two years in a row. With the time that became available, colleagues were able to make it impossible to make the wrong transfers. That is exactly what we want to achieve: now colleagues can focus on improving and making their work more efficient. But not everyone immediately sees how RPA can help. That’s why it’s my job to convince them. Fortunately, I am now able to share quite a few success stories, which helps. In addition, one of my biggest challenges is to show my colleagues that robots are not a risk. They are often afraid that robots will damage their system or make mistakes. I explain: robots are exactly what you control, they do nothing more than execute the script you make yourself. That’s actually safer than asking a human to do that work; you know exactly what robots do and don’t do.

If people are convinced, one of the six analysts in my team will help determine how a Finance department can work more efficiently and what role RPA plays in this. The ten IT developers then build the bots and four support engineers ensure, among other things, that the bots have a virtual workplace, including login names and passwords for access to specific systems’.

Better every day

I never expected to work with RPA myself. I have a history as an SAP consultant and, for example, I can’t program, but I do understand the possibilities of RPA. Within our team we also accept that new people still have a lot to learn. There is no other way, RPA is still too new. RPA suits me because I want to do things better every day. I am always looking for ways to work more lean. Recently I also got my Lean Black Belt. One more step to the master black belt and then I reached the highest achievable level of lean working. I now see robots as an indispensable way to set up departments more efficiently’.

Meanwhile, our first robots have retired
Rianne Selhorst

200,000 hours of robot work

“I couldn’t have imagined we’d achieved so much in three years. Our bots within Finance have already done 200,000 hours of work in the past three years, which has significantly reduced the workload of colleagues. Of course, my team wants to achieve a lot of new things with our robots in the coming period. For example, we are now developing a chatbot to answer questions about invoices. If this works, we also want to be able to use the chatbot to give assignments to our RPA robots. Now colleagues still do that by asking our support engineers. We are also investigating whether we can apply OCR technology to the bots. This allows computers to read text from an image or PDF. The robots would then also be able to read signatures and handwritten texts. And, of course, we will continue to look at what work can be robotised, including our own. Our team already uses a support robot, Sporty. It resets passwords of other robots, clicks away pop-ups and collects and reports error messages. And of course we hope our bots can retire soon thanks to their hard work. Who knows, maybe Harry will be next.’