RPA (Robotic Process Automation) has proven to be effective in automating tedious and repetitive processes for traditional IT environments. The ROI (Return on Investment) has generally been quick and the technology can be a stepping stone to other modernization efforts, such as with AI.
So in light of all this, it is no surprise that the RPA industry has seen substantial growth. Just look at UiPath, which recently came public and has a market capitalization of $32 billion.
Note that the ARR (Annual Recurring Revenues) is running at $653 million, up 64%. The dollar-based net retention rate is 145% and there are over 8,500 customers.
Now when it comes to RPA, the use cases are often about automating Windows applications. But there is another interesting category—which is often overlooked—that can provide for growth opportunities: mainframes.
Of course, the perception is that this market is a backwater. Yet this is really a myth. Keep in mind that mainframes power many of the mission-critical workloads for the world’s largest enterprises. For example, 92 of the world’s top 100 banks use mainframes as well as all of the top insurers and 18 of the top 25 retailers.
Granted, it’s true that mainframes need to be modernized and these efforts have been expensive and time consuming. So can RPA help out? Well, interesting enough, it has already done so.
“Many early exemplar use cases leverage accessing mainframe data and transferring it to and from other systems of record like SAP, Oracle, and so on,” said Ben Chance, who is the Vice President of Intelligent Automation at Genpact. “The large majority of these use cases were in the business process automation space, spanning financial services and other large legacy footprint organizations.”
However, mainframes pose tough issues with integration. There are often a limited number of APIs. What’s more, RPA usually interacts with mainframes with screen scraping, which can be challenging because of multi-screen configurations and complex processes.
“Organizations should consider using attended bots–which means the end user would execute the bot by clicking a button–to augment the manual process,” said Bob Grabowski, who is a Managing Director at Deloitte Consulting. “Imagine a scenario where a human performs the first few steps of a manual process, then kicks off an attended bot to complete the next 15 to 20 steps in a matter of seconds, then the human finishes the process with several more manual steps. These attended bots may not automate the end-to-end process, nor will they have widespread use across the enterprise, but they can be very effective tools to accelerate several manual steps in a process that is executed thousands of times.”
Another issue is with transactions. Mainframes often use sophisticated applications like CICS that handle huge workloads, say for processing millions of records.
“Pure screen scraping approaches are highly inefficient and tend not to scale well in many environments,” said Dr. Alex Heublein, who is the President of Adaptigent.
But there are ways to solve the problem, as seen with software from HostBridge Technology. Founded 20 years ago, the company is focused on developing soluitions that allows for integration and orchestration with IBM mainframes.
“Lately, our customers are using HB.js to write RESTful services that their UiPath or other bots and automations can call,” said Russ Teubner, who is the CEO and cofounder of HostBridge Technology. “Integrating bots to the mainframe in this way provides a scalable, high-performance integration path that lets you avoid the perils of screen-scraping.”
For the most part, the opportunity for integration with mainframe environments is enormous. In fact, the COVID-19 pandemic has created more urgency for this. And while RPA can be helpful, there is definitely more room for innovation.