Intelligent Automation vs Hyperautomation – what’s the score?


One of the challenges we face as analysts is how to categorize technologies. I’m feeling this directly as I work through the (still-evolving) area of automation platforms, which are coming out of the convergence of robotic process automation (RPA), business process management (BPM), intelligent document processing (IDP) and the increasingly ubiquitous machine learning (ML), including natural language processing (NLP).

This convergence is clear, caused in part by the “law of diminishing returns” for the individual categories. For example, an enterprise may procure an RPA solution to cover a broader set of processes, but deployment can hit limits due to factors such as process complexity and scope (whether processes run across multiple teams or departments). This drives a need for more powerful solutions with broader functionality. 

While the pressure is on for vendors to broaden what they offer, this is an evolving area where no single vendor has completely defined each piece. Multiple terms are becoming prevalent: intelligent automation (IA), intelligent process automation (IPA), hyperautomation platforms, and intelligent automation platforms (IAP). Confused yet? Let’s triage the terms – noting that each, in essence, is looking to offer a “one-stop shop” for automation end to end. 

First, IA is not referring to a platform but to the idea of using AI to augment the solution and help in decision-making across the enterprise. It can refer to IDP, RPA or BPM, all augmented with AI. Drilling into this, IPA is the next-gen RPA on steroids, leveraging AI/ML/NLP. While this focuses on RPA, many vendors are adding IDP as a part of this so that the processes can start with unstructured documents to automate and process.

In researching, very few vendors call their solutions hyperautomation platforms. They discuss hyperautomation as a strategy to intelligently automate as many processes as possible in an enterprise, using a variety of tools to automate end-to-end. So, you may have an IDP solution, a different RPA solution and a BPM solution with AI, all working together, versus a single solution providing all or most of the end-to-end automation. 

And what about IAP? From a research and briefings perspective, I’ve found a very small number of vendors with all the pieces – IDP, RPA, BPM and AI/ML/NLP. Many are strong with two-thirds of the picture, such as IDP/RPA or BPM/process discovery. These solutions often tightly integrate with other industry leaders for the pieces they are missing. 

We’ve currently settled on IAP for our research, as this appears to be where the puck is going – platforms that support all the capabilities listed as an integrated whole. We chose to focus on the IAP rather than hyperautomation, because the latter focuses on a company’s overall IT strategy and disciplines to run the business, whereas IAP drills down on the actual technology and solution to support this. 

As a strategy, hyperautomation brings end-to-end thinking into the mix. An overarching vision is wise, and all organizations should have the big picture in mind to maximize efficiency. However, this should not be a barrier to entry: you can adopt an IAP without an enterprise-wide strategy. Many companies start smaller, and the broader strategy can grow from smaller experiences and successes.

Overall, most vendors I speak to are describing hyperautomation as a strategy that their product can help with. They are also heavily pushing the idea of an intelligent automation platform that facilitates end-to-end automation with ease of use and no code/low code, security, compliance, etc., all in one platform. A platform that helps a company do things right. 

In summary, this isn’t an easy topic to scope, and it is still rapidly evolving. We’ve settled on IAP based on our view of the market right now; meanwhile, while you can have IAP without hyperautomation, if you are smart, you will consider the former within an overall vision and strategy for the latter. 



Latest articles

spot_imgspot_img

Related articles

Leave a reply

Please enter your comment!
Please enter your name here

spot_imgspot_img