UiPath – From RPA to Hyperautomation
More than 75% of people think they’re above-average drivers. Psychologists refer to this as “illusory superiority,” and it’s the propensity for all of us to overestimate our good qualities and capabilities. In the workplace, people are even more motivated to exaggerate the positive impact they’re having on the entire company. This can pose a problem during process mining.
In a previous article, we wrote about process mining vs. business process discovery. Both of these terms refer to the act of defining repeatable business processes. You may have heard about robotic process automation (RPA) which is what naturally happens after you define a repeatable business process – you automate it using artificial intelligence (AI). Today, we want to talk about a company that is now building on top of RPA with something called “hyperautomation.”
Founded in 2005, New Yawk startup UiPath has taken in a whopping $1.2 billion in funding from names like T. Rowe Price, Tencent, and Sequoia. Last month’s Series E of $225 million gives UiPath a valuation of just over $10 billion. All that funding is being used to propel the fast growing startup into a leadership position at the top of the 10 RPA vendors that control 70% of the market today. The below Magic Quadrant from Gartner shows UiPath leading 14 other vendors in respect to execution and completeness of offering:
In our recent piece on Making IBM Great Again With Artificial Intelligence, we noted that larger companies are now making smaller acquisitions in an attempt to compete with the ‘big three’ RPA vendors – Automation Anywhere, Blue Prism, and UiPath. And it’s not easy to get it right. As many as 30 to 50% of initial RPA projects fail, but those that don’t are reaping huge rewards for companies across the globe.
What is Hyperautomation?
Hyperautomation – also sometimes referred to as hyper automation – is explained by UiPath as something that happens after successful process automation:
It starts with robotic process automation (RPA) at its core, and expands automation capability with artificial intelligence (AI), process mining, analytics, and other advanced tools. The idea is to automate more and more knowledge work, and engage everyone in an organization to be part of the transformation.
We can try to visualize what this might look like using the below example.
An Example of Hyperautomation
Anyone who has served time in a large multinational corporation knows about “that report.” It will be some collection of obscure information that’s vital for an entire division to function properly, but only one person knows how to produce it. It’s usually the result of a skunkworks project by some senior BSD who got tired of waiting for IT to get around to creating reports and decided to ask for forgiveness instead of permission. With hyperautomation, applications are now expressed as digital workers with whom you can interact with and ask questions. Creating a method for anyone to easily access any information they’re permissioned to access lets everyone find out what they need to operate at peak efficiency.
The next step after automating business processes is to allow anyone to interact with them using natural language. And we’re not talking about a cheesy “ask me anything” Salesforce search box which falls flat on it’s face if you ask it questions like:
- @Salesforce Can you find out which customer service person worked on the most tickets for CLIENT X last year? Then please ask @HR how long they have worked for us and how big of a bonus we gave them last year?
In the above example, the person asking the question is able to engage with an application and a department in order to find out information – some of which is sensitive – that can be used to inform a senior manager during compensation time. For sales managers, being able to ask Salesforce or LinkedIn questions in natural language empowers them to make more informed decisions.
- @ Salesforce What is the run rate for all accounts SALESPERSON X is assigned to and please exclude CLIENT X?
- @LinkedIn, I’m going to see JOHN DOE at CLIENT X. Can you please tell me how long he was been working there and what he does?
If a chatbot can understand natural language perfectly 99% of the time, then just about anyone should be able to start asking questions to enterprise applications or functional departments. Think of it as your own internal Google. Anyone who is permissioned to do so can now ask questions of the data without having to go through some drawn out process to get it. This is the future of hyperautomation.
Several months ago, UiPath published a press release titled UiPath Brings Conversational AI Capabilities to Industry’s First Hyperautomation Platform which talks about how they’re partnering with a firm called Druid to offer chatbot interfaces the are both internal and external facing. Natural language processing has now reached a maturity level such Druid is able to “support over 40 languages, countless internal and external channels, and offers more than 300 pre-built conversational AI templates covering business scenarios across multiple industries and roles.” Says the release:
Not only are agents’ workloads reduced by customer-facing bots who handle repeated questions over messaging channels, but agents are further supported by bots after escalation.Credit: UiPath Press Release
Today’s pandemic has accelerated the demand for customer service with many institutions telling their customers to expect deteriorated levels of service. That’s where UiPath can help. Bank Romania implemented an automation solution that integrates the capabilities of both Druid chatbots and UiPath’s software robots to process requests to postpone bank loan installments. The integration allowed the bank to cut down the processing time of a single request from 10 minutes to 20 seconds, and cope with a 125% increase in the number of calls received by call center agents, enabling it to process three times more deferral requests with the same number of people in the back office. Every manager is now being asked to do more with less, and this is a good example of why RPA is now the fastest growing category of enterprise software.
Hyperautomation is about extending RPA beyond the back office and into client facing situations. Any client should be able to change their address in your CRM system, but imagine all the process that goes into making such a change. Now, chatbots handle all incoming requests as a rule and bring in “humans in the loop” as needed.
How UiPath Stacks Up
The Garnet Magic Quadrant for RPA vendors we looked at earlier analyzes each of the vendors it lists along with strengths and weaknesses. For UiPath, strengths include the ability to integrate with “most all major enterprise products and applications,” along with the resources needed to accelerate implementations such as
A community of more than 750,000 developers can share automation and AI libraries in its marketplace, from which they are available as prebuilt, reusable components.Credit: Gartner
In other words, they’re helping people help themselves. Under weaknesses, Gartner notes “product pricing complexity” and the following growing pain:
Reference customers of UiPath observed that UiPath’s rapid growth has resulted in it losing some of its customer touch as demonstrated by its disconnected operations and lack of coordination between software and service departments.Credit: Gartner
Sounds like UiPath isn’t eating enough of its own dog food.
In the blue collar world you have cobots, and in the white collar world you have hyperautomation – people working alongside software robots. Instead of programmers figuring out what to automate you now have citizen developers. Everyone helps figure out what to improve. It’s that “constant improvement” that the innovation management experts talk about. Some of the success stories sound too good to be true. In a press release announcing the release of their hyperautomation platform, UiPath mentioned the following success story:
One of the world’s largest professional services firms has championed A Robot for Every Person and upskilled their 55,000+ employees with UiPath to enable their employees to deliver world-class insights to their clients. By using a citizen-led approach that invites employees closest to the processes to build their own automations, this initiative has seen several million staff-hours saved and annual savings of more than half a billion dollars.Credit: UiPAth
Just to emphasize that last part – annual cost savings of more than a half a billion dollars. Says the Chief Product Officer at UiPath:
We estimate that for $1 invested in UiPath, our customers can return $15 or more in their first year.Credit: UiPath Press Release
It’s no surprise then that over 50% of the Fortune 500 are now using the UiPath platform to automate complex processes and reduce operational costs.
It’s tough to avoid some of the buzzword bingo that accompanies hyperautomation. Look no further than a piece Gartner published titled “RPA Renaissance Driven by Morphing Offerings and Zeal for Operational Excellence,” which makes hyperautomation sound more like the second coming of Christ. The truth is that this “productivity revolution” is simply companies like UiPath helping other companies learn to leverage machine learning to do things more efficiently, something that ultimately affects the bottom line.
Despite what the pundits say, FAANG stocks (Facebook, Apple, Amazon, Netflix, Google) don't give you real exposure to AI. Read about 7 stocks that give you true pure-play exposure to AI in our guide to investing in AI healthcare companies, freely available to Nanalyze Premium subscribers.