Optimizing robotic process automation
Advancements in robotic process automation are enabling companies to quickly introduce improvements to outdated, complex, or error-ridden processes. But it’s not as simple as setting a bot loose and seeing massive returns in time and money saved.
At ATB we are very strategic in how we are rolling out robotic process automation. From how we decide which processes to tackle, to rolling out pilots, to tweaking and optimizing the processes themselves, we’re putting great thought into how we can leverage robotic process automation to create better experiences for ATB customers.
Our first priority is to identify where we can create the most value or solve the most pressing customer problem. With over three thousand internal and customer facing processes at ATB, we had to devise a strategy to evaluate and prioritize where to leverage automation.
Guided by our North Star, our mission, of zero-disappointed customers, we started by documenting as many of our processes as possible to identify the right opportunities. This list of 1,700 process was then layered with factors like the risk level, process volume, complaint frequency, suitability for automation, and level of importance for customers.
The result is a heat-map and listing of our internal processes by order of priority for automation.
One of the processes that came out on top was customers making changes to their address on file. An occurrence that comes up frequency in the the lives of ATB customers. Internally, we have multiple systems in which an address change needs to occur, with plenty of downstream effects if there are delays or errors. For a customer, the request is straightforward: ‘I’ve moved, I need to change my address.’ And, in the middle of a move, they have no time to spare and don’t need any extra stress! This combination of factors makes address changes a prime candidate for a robotic process automation solution.
One of the other things that our process automation team does in our evaluation of processes is to take the lens of ‘do we need to do this?’ and ‘does this add value for the customer?’ We don’t want to automate bad processes. We want to make necessary processes smarter, improve the business rules, and streamline the steps in a process.
Automation and process optimization are not new concepts. What is new today is the ease with which they can be applied and the broad use cases and scale of solutions. Where backend integrations usually require the accommodation of every possibility, robotics can isolate outliers that don’t pass validation checks or adhere to business rules, breaking the ‘all-or-nothing’ paradigm of automation.
As we build our internal capabilities, the applications of drag-and-drop solutions and robots-as-a-service are enormous. Where connections with third party applications may be limited by permission to make changes or batch processes, robotics can allow us to connect even to external programs through APIs. Our early applications of robotic process automation are also scalable and adaptable to other business areas with little to no development work required.
We are still early in our journey with automation. When we started, we used robotics software that mimics human actions. It basically automated the repetitive task of entering information. It picked up the request in one location and actioned it across multiple systems based on predetermined rules. For example, a staff member changes an address in a certain location and it activates an automated process.
Next, we added an app script to improve the user experience of initiating the request and remove the barrier of knowing in advance what form to fill out to activate the process. The third evolution is to layer on machine learning and natural language processing to allow for dynamic interaction.
Robotic automation is what will ultimately bring the capacities of artificial intelligence and machine learning to customers. As the team progressively learns and improves, we are moving from local automation of one-off processes to creating end-to-end solutions.
We are moving towards turning natural language inputs into structured data that results in action, rather than an information response. So you say ‘I need to change my address’ and, instead of written instructions about where to go and what to do, a responding information request actions an automated backend process.
We’re focussed on applying technology in areas that impact customers most. Robotics process automation applied to the right use-cases means better customer experience, and more time for us to hear our customers. Just one of the ways we are working every day to make banking better.
Are you passionate about implementing technology to solve customer problems and improving customer experiences? We’d love to hear from you, check out open positions on the team.