CTO of the Corporates Customer Segment Thomson Reuters
Share this content

Not all automation is the same: The difference between RPA and API

7th Aug 2019
CTO of the Corporates Customer Segment Thomson Reuters
Share this content

Shawn Malhotra outlines why companies need to understand automation to meet their financial, tax, and compliance requirements and explains the difference between Robotic Process Automation and an Application Program Interface.

Like it or not, automation is now a key component to digitally transforming the way organisations operate. But is it a good or bad thing for businesses?

Sometimes referred to as Business Process Automation (BPA) or digital transformation, it is the technology-enabled automation of complex business processes that can help simplify manual processes, increase service quality and improving delivery while, ideally, containing costs.

Companies are now moving beyond the hype surrounding automation and realising that a level of automation in some shape or form brings a plethora of business benefits to their operations.

However, like all technological advancements, automation is evolving, and many may not clearly appreciate that not all automation is the same.

From a financial, tax and compliance point of view, business leaders and heads of departments will likely be looking at either adopting a Robotic Process Automation (RPA) or an Application Program Interface (API) solution to meet their requirements.

It is at this point that they need to clearly understand that there is a difference and it is one that many miss in their urgency to automate.

Put simply, both an RPA and API will remove the human from some or all of a particular process. Both either automatically pulling information from a system, processing it in some way and injecting it back to a (possibly different) system.

You can think of RPA as a software robot designed to mimic human actions to complete a specific task, whereas APIs integrate across applications as 'digital' plumbing, allowing programs to interact with each other.

The problem with RPAs is that they are generally harder to maintain. The RPA in question assumes that the interface it’s interacting with is a user interface and it makes an assumption based on the user behaviours today, but not in the future.

It doesn’t have the flexibility or agility to adapt. An RPA will, therefore, tend to cost more and require more maintenance to ensure it doesn’t break.

On the other hand, an API is designed to interact with machines. It’s still achieving the same thing as an RPA, only better. It is a well-defined software-to-software interface, not a user one. It means the applications talk to each other without any user knowledge or intervention and it can cope with huge amounts of data compared to an RPA.

For example, if a business operates across multiple jurisdictions it will have to cope with a variety of foreign languages and information. A simple data point such as the date relating to a piece of tax information filed will be presented in different formats.

An RPA will struggle to process these different formats and potentially as different countries come online the RPA will have to be updated to cope with their different formats. With an API however, the user inputs the data into their system, no matter the language, the API interacts with this and pulls back the date (and data) without the user having to ask the question differently and presents the required information back. An API removes the complexity completely.

The challenge is that when businesses are looking at automation they may not be aware that an API is the best route. Many tend to ask for an RPA from providers and can find themselves with an automation solution that’s not fit for purpose.

Instead, they need to ask providers whether they have an RPA or API available and if they only have an RPA, whether that’s a stop-gap solution while they develop an API. It’s not a mistake, but RPAs are the more costly route to automation as they require maintenance, tend to break and require debugging.

Being designed for machines rather than humans, APIs will deliver greater bandwidth and will be able to cope with future requirements as they are updated centrally by the API owner.

The machines are not replacing humans. They’re helping them by delivering more accurate data and accelerating processes that will help business leaders to make better-informed decisions, in a more timely and cost-effective manner. As such people need to understand what they’re asking for and make the right choice of automation technology to avoid disappointment.

Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.