Deploying RPA: why DevOps and IT need more control

In the old days, IT was the backbone of a business’s digital infrastructure. If a business wanted to implement new technology, IT had to plug it in and run it for the business to use.

But then the SaaS market exploded. Suddenly, non-IT departments could buy tech and implement it themselves and a bunch of CEOs starting walking around like they were Mark Zuckerberg. This trend became known as ‘shadow IT‘ and while cloud computing, SaaS and the bring-your-own-device (BYOD) phenomenon all played their part in its rise, arguably, it was robotic process automation (RPA) that opened the floodgates.

RPA is a type of software that automates rudimentary rule-based processes. By interacting with application interfaces just as a human would, RPA software can carry out all sorts of tasks such as data entry and invoice processing. RPA tools are also typically technology agnostic; meaning they can work across a whole host of IT platforms such as ERPs, CRMs and custom applications. This means enterprises have an alternative to replacing legacy systems and there’s no need for connecting to them via APIs, all of which costs precious time and money.

Furthermore, because RPA has been commonly sold to enterprises under slogans such as ‘no-code’ and ‘easy to integrate’; and marketed to non-IT departments, many enterprises have begun considering it as an alternative to IT and DevOps.

According to Ollie O’Donoghue, research VP, IT Services at HFS Research, this push for RPA has also been driven by a general feeling among enterprises that IT is just not doing enough.

“Non-IT departments have targets and ambitions to transform their business and feel frustrated that IT is just trying to keep the lights on,” he said. “So when a technology like RPA comes along and it’s pitched and marketed to a business audience and they can see positive results almost immediately, it’s a no brainer for them that they’re just going to try and run it themselves; rather than have a lengthy conversation with IT over how to best implement it or how it fits within their technology roadmap.”

Is RPA a credible alternative to DevOps?

But should DevOps be worried? According to O’Donoghue, no. He said: “Ultimately, RPA does not take away the bulk of what DevOps and IT services teams do. There’s a whole spectrum of tasks their busy with from on-the-spot patching and service development. RPA can only do a very tiny part of this. So we’re never going to see a direct competition between RPA and DevOps, which is more of a cultural methodology for IT development and operations.

However, because RPA can be introduced by not-IT depatments a whole host of challenges have arisen for DevOps and IT,  argued O’Donoughue.

RPA’s a sticking plaster

According to him, when non-IT departments start introducing RPA to their enterprise, they tend not to integrate it with the wider IT infrastructure. In many cases, they’re just extending the life of an old legacy system, which means there, essentially,  just buying themselves time. In other words, RPA introduced under these conditions will act as a sticking plaster.

“This business-led approach is fine for integrating stuff that you cannot be bothered to upgrade but when you get into the complicated processes and the heavy lifting, you actually need developers involved because they understand the services ecosystem and they have advanced skill sets that are required,” he said.

Don’t be an RPA tourist, implement effective change management

The current approach taken by many enterprises to introduce robotic process automation is broken; it’s time enterprises paid more attention to their people and implement effective change management

According to a survey by HFS research, of close to 600 major global enterprises, only 13% of RPA adopters are currently scaled up and industrialised and they succeeded because they introduced RPA as part of an integrated strategy.

Speaking with Information Age, Phil Fersht, CEO, HFS Research, said: “Beyond scripts and bots and dreams of digital workers scaling up rapidly to provide reams of value, most enterprises are fast coming to the realisation that they need an actual process automation platform capability that ingests their data, visualises it, machine learns it, contextualises it and finally automates it.

“Essentially, the whole lifecycle of data components needs to be integrated into a single platform in order to take maximum advantage out of automating processes through scripts, bots and APIs.”

Scaling RPA: before automating processes, improve them

Most enterprises aren’t scaling RPA across their entire organisation, in part, because they don’t understand their processes to begin with. Featuring more insights from HFS Research

Why DevOps and IT need to get involved in RPA initiatives more

The disconnect between business and IT must stop. As HFS Research’s analysis points out, while RPA is appealing to business-centric people, it requires buy-in from people with strong technical knowledge to truly succeed.

DevOps can play a role in helping deploy RPA properly, they can ensure RPA deployments are closely aligned and integrated with a company’s strategic objectives.

Without a clear path, enterprises can fall into the trap of automating for the sake of automating, without taking practical and technical objectives into consideration.

Burnett: Look beyond RPA hype

Despite talk that RPA is overhyped, Sarah Burnett, a guru on automation technologies, says that it can cut process costs by 30%, but there is another benefit, not so obvious, and it lies with creating data we can trust

Avatar photo

Andrew Ross

As a reporter with Information Age, Andrew Ross writes articles for technology leaders; helping them manage business critical issues both for today and in the future