APIs and robotic process automation can be considered competitive approaches for driving automation in the enterprise. Each approach has its own merits and drawbacks. However, CIOs can simplify their automation strategy by focusing on a common automation platform for building new automation rather than a specific technique.
"In a fight against system bloat, I would suggest your RPA engine, workflow platform and business rules engine can and should be one and the same or driven as one consistent layer and strategy," said John Cottongim, automation director at Mars, the candy giant.
His focus is on creating a flexible, no-/low-code business rules engine that works across all automation scenarios. "For the time being, this is where a robust automation platform that enables API and UI automation, while delivering rapid development of business rules and decisions methods, shows its value as an enterprise platform," he said. Whether there is a platform delivering best-in-class performance across each of these categories is yet to be seen, though it is certainly not hard to imagine.
Finding harmony between RPA platforms and APIs
The capabilities of both RPA platforms and APIs are evolving to support use cases primarily handled by the others. The combination of RPA and APIs is a natural outgrowth of the modern business systems environment, particularly driven by the adoption of SaaS platforms and API-first becoming the new software mantra.
Traditionally, RPA has been marketed to work with the complex mix of legacy, third-party and modern business applications that most organizations have accumulated. When delivering an RPA platform, it is nearly always best to use APIs when available, as the combination of these technologies delivers an extensive and change-resistant experience by removing the inherent change-prone UI layer from the equation.
"Counter to what some may assume, the existence of an API does not negate the usefulness of RPA," Cottongim said. In an API-rich environment, RPA shows its value as a no-/low-code business decision and workflow orchestration platform. It can also be an effective technology to deliver quick value with small investments that connect gaps in a changing technology landscape or business process.
Cottongim doesn't see great improvements in either of these technologies per se, though the pervasiveness and standardization in implementation certainly has helped improve adoption and speed to value, particularly for APIs. Ultimately, APIs are a great tool to ingest and send requests from or to a system. However, they are not able to effectively construct unique and complex business processes.
RPA platforms are about workflows
"RPA has become more about automating workflows end to end and not just about connecting two systems," said Goutam Nadella, executive vice president of client solutions at Symphony, a financial service messaging provider. In the past, enterprises focused on connecting systems to each other with RPA. But, today, he sees RPA as being more about bringing humans, bots and applications together to enable better workflows from an end-to-end perspective.
Nadella said that, when Symphony's banking customers are talking about RPA, they are thinking about a workflow and all the steps a user is taking and then identifying how many of those steps can be repeatable. From there, they can use APIs to execute one or two of those steps because APIs are just one component in the bigger RPA picture.
In essence, the bots are either operating with no input from humans or fetching information in collaboration with humans. Either way, the bots enable humans to be more efficient. But, underneath the services, the underlying systems need to have APIs. "APIs are a foundational layer to any system to have bots to exist and to function well," Nadella said.
Competing for CIO mind share
In the short run, both API management vendors and RPA platforms are making inroads into each other's territory, with the goal of becoming the automation platform of choice. "CIOs should approach automation deployments holistically, with both APIs or RPA serving an integral role in the process," said Steve Shah, vice president of product management and enterprise at Automation Anywhere.
Part of the uptick in the recent enthusiasm for RPA may lie in its ability to bridge the worlds of traditional APIs and a new generation of AI capabilities. "By simplifying the integration of AI, RPA can help automate even more complex tasks that involve both structured and unstructured data, granting anyone the ability to create turnkey solutions to previously tedious tasks requiring human intervention," Shah said.
According to Shah, drag-and-drop UIs for developing automations could be a game changer. This much simpler UX for business app development means that those who understand the business process can easily create their own automation and drive greater efficiencies.
API vendors unsurprisingly believe that API management tooling will become the platform of choice. Ross Garrett, chief product officer at Cloud Elements, an API vendor, said that RPA is just the latest buzzword in the world of integration, such as the no-/low-code movement and various others, like computer-aided software engineering dating back as far as the '80s and '90s.
Garrett acknowledged that improvements in AI are making it easier to automate more tasks, but the rigors of API management are required for proper governance and security. "Ultimately, API-based integration is the enabling layer -- the foundation that supports modern enterprise integration and makes new ideas like RPA possible," Garrett said.
In the meantime, both RPA and API management vendors need to establish themselves in the minds of CIOs before the large cloud players become the front end for both APIs and RPA. Sebastian Dolber, CEO of Astor, a cloud solutions consultancy, sees big enterprise software vendors, like Microsoft and Amazon, closing in on specialized RPA vendors by offering alternatives, such as Microsoft Azure Logic Apps, or Amazon Simple Workflow Service. These offerings "seem to hit the spot on flexibility between classic RPA and APIs," he said.