Artificial Intelligence in Business Process Automation
“I can’t wait to push some paper today!” Said no one ever. The mind-numbing work to keep the wheels of commerce rolling—filling out invoices, deciphering hand-written memos, processing insurance claims—can be a real grind. It’s been that way since the time when Ebenezer Scrooge refused to provide another lump of coal to help warm overworked clerk Bob Cratchit. Lacking frailty of mind and body, artificial intelligence for business process automation appears to be a no-brainer.
In fact, a number of companies are employing AI techniques such as machine learning, computer vision and natural language processing to automate business processes. White-collar outsourcing is no longer going to India—it’s moving to the cloud. On one hand, that means job losses in the short term. It’s already happening. The tech-loving Japanese are among the first to be replaced. For example, IBM Watson is now doing the job of more than 30 employees at an insurance company by calculating payouts. On the other hand, there’s a lot of money in outsourcing. Global business process outsourcing was worth $63.5 billion in 2015, according to Statista. Companies rolling out artificial intelligence for business processes say automation will free employees from mundane tasks for more dynamic work, like checking Facebook more often.
You may recall that a few months ago we expressed some healthy skepticism about this new, so-called digital workforce. Our complaint was that much of this seemed like software automation repackaged with buzzwords like “robotic process automation,” “cognitive technologies” and “desktop automation.” However, we promised to dive further into this sector. In this article, we’ll look at a couple of startups and a couple of veteran companies claiming to use artificial intelligence for business process automation.
Artificial Intelligence in Business Process Automation
WorkFusion – AI-Powered Bots
New York-based WorkFusion is fresh off a $35 million Series D in January 2017. That brings the total investments to about $71 million since the company was founded in 2010. The latest series was led by Georgian Partners with participation from existing investors Mohr Davidow Ventures, iNovia, Nokia Growth Partners, Greycroft and RTP Ventures.
WorkFusion claims its machine learning platform can eliminate up to 90 percent of back-office business work and its AI-powered bots can increase service center capacity fivefold. Clients can reportedly enjoy a 50 to 80 percent ROI within the first year.
How does WorkFusion work its magic? Remember that machine learning is all about improving computer performance through experience rather than pre-programmed software. Its AI platform learns on the job. It has already studied the habits of 35 million users on its platform on how to do business-related processes, according to a story in Business Insider.
Last year, WorkFusion partnered with VirtusaPolaris, the “market-facing brand” created when Virtusa Corporation (NASDAQ: VRTU) acquired Polaris Consulting & Services a year or so ago. VirtusaPolaris provides IT consulting and outsourcing support in the banking and financial services market. That would seem to open up bigger opportunities for fledgling WorkFusion. Though we wonder where this partnership might eventually go. Virtusa has been on a shopping spree since 2009, acquiring seven companies during that time.
HyperScience – Reading Documents Faster
Another startup from the Big Apple using artificial intelligence for business process automation is HyperScience. It comes to the table with nearly $19 million in funding, most recently a second Series A in December 2016 that netted $8 million, led by Felicis Ventures. That’s also when it announced its existence to the world with the launch of its website.
The company initially focused on back-office automation. Its first product, HS Forms, is meant to replace the tedium of data entry. Its machines can read and understand any kind of text, apparently even our atrocious scribble. Such automation speeds up the processing of products like mortgage applications or medical records faster than any human, even with a double espresso each morning. HyperScience says HS Forms uses “advanced computer vision techniques to process documents, identify content types, and extract the content.”
The company also lists two additional, more sophisticated products. HS Freeform can take unstructured data, whether digital or handwritten, and read, understand and digitize the information. HS Evaluate goes one step further, reviewing files and applications with human-like judgment. The company says its AI software “can automatically review an extensive claim file, eliminate duplicate entries, assess eligibility, and then deliver precise adjudication decisions.” Judge Judy, you’re fired.
IPsoft – Cognitive Agents
Apparently the water in New York isn’t just good for making bagels. Yet another company applying AI to business process automation is IPsoft, a private company with offices in 11 countries. It’s been around since 1998 as an “autonomic and cognitive solutions service provider,” but recently went full AI with Amelia. Amelia isn’t just another chatbot, according to the company, but an artificial intelligence platform that can automate just about any business process currently done by bipedal cubicle critter. Amelia’s digital job resume includes everything from helping customers open new bank accounts to processing insurance claims.
Calling her a cognitive agent, IPsoft says Amelia can emulate human intelligence, making her capable of natural interactions with people. “She can understand human language, learn through observation and determine what actions to take in order to fulfill a request or solve a problem.”
In one case study, for example, Amelia was able to take over nearly 20 percent of all incoming IT service desk tasking for a European bank after only 45 days of training. And the multi-tasking cognitive agent proved to be a fast learner as a mortgage broker as well. Within two weeks, Amelia could answer three-quarters of all questions with an 88 percent success rate.
While we’re quite skeptical about chatbots, it looks like Amelia is where AI finally begins to add some value via a chatbot interface. What a nice innocuous name, Amelia.
Blue Prism – Robotic Process Automation
From across the Pond we found publicly traded Blue Prism (PRSM.L). The U.K.-based company, with offices across three time zones in the United States and Australia, debuted on the AIM market of the London Stock Exchange in March 2016. So far, the company—at least its stock—has performed admirably. An investment of £10,000 at the company’s first closing bell on March 16 would have netted you nearly £40,000 less than a year later.
Founded back in 2001 by a group of process automation experts, Blue Prism counts more than 150 enterprise clients, including ten top global banks such as Barclays Africa Group, BNY Mellon, Commerzbank, Nordea, ING and Westpac, as well as several of the world’s leading insurers including Zurich, Swinton Insurance and Aegon. Other big names in various industries include Maersk, Siemens, IBM, Procter & Gamble and Nokia.
Here’s how Blue Prism explains its robotic process automation (RPA) platform:
Blue Prism’s RPA is built as the transactional platform, with its software robots helping AI turn decisions into actions. The cognitive decision making built into these software robots brings AI to life and also enables knowledge transfer between the robots and third-party AI applications as the two engage and interface with one another allowing the robots to recognize items and take action with no external intervention.
We see quite a bit of momentum in AI-powered business process automation. More than a few Global 2000 companies are signing up with players like Blue Prism and IPsoft. It won’t be long before you’re conversing with Amelia rather than “John” in Mumbai.
Conclusions on AI and BPA
In a sense, this sort of automation has been around for a while, as we noted earlier. It’s the artificial intelligence piece that’s new. And like with many of the pieces we write on this topic, we want to caution you that anyone can slap AI on their website and call it machine learning. It’s like seeing the word “natural” on a food label. You have to be sure to read the ingredient list first.
If you enjoyed this article, then sign up for our free newsletter - Nanalyze Weekly. About every week, we'll send you a simple summary of all our new articles. If you didn't enjoy this article, share it on Twitter and tell everyone how much you hated it.