In this third year of CFO’s Tech Companies to Watch, we found a notable shift in the applications being offered by vendors. In every functional area of the companies that made our annual list — process management, recruiting, travel expenses, receivables management, and compliance among them — the goal is unearthing the real story behind the data to make more intelligent decisions.
We can’t tell you how much heavy lifting is needed to get a real return on these products. However, the profiles will help you gauge whether these tools deserve a closer look by your organization as it journeys toward a smarter version of itself.
The entirety of the list appears in the April/May 2019 edition of CFO magazine. We are revealing the first 10 companies on CFO.com, one per day, beginning April 24.
Guiding RPA Investments
Many traditional finance activities, as well as other corporate processes, can be automated. But executing the automation is just the back end of a much bigger task: deciding which processes are the best candidates.
Robotic process automation (RPA) vendors, for example, are better at building bots to take over particular processes than they are at helping companies tackle that bigger task. That’s where a growing category of artificial intelligence companies like FortressIQ comes in.
FortressIQ emerged from stealth mode in December 2018 with $16 million in venture funding. As of February, CEO Pankaj Chowdhry told CFO, the company had 6 enterprise clients, as well as pilots with 3 of the world’s 10 largest banks and one of the largest consumer product companies. He added that FortressIQ was ready to sign contracts with more than two dozen customers in the second quarter.
What’s all the excitement about?
Picture a huge company with thousands of back-office workers who regularly perform tens of thousands of processes across finance, accounting, procurement, human resources, and other functions. It has a $2 million budget for process automation this year — but what exactly should that money be spent on?
The company could hire a big consulting firm to come in, disruptively interview hundreds of staffers, watch them work, and ultimately recommend the best candidates for RPA. Alternatively, it could plug in FortressIQ’s cognitive process analysis software and achieve the same result at a fraction of the cost and in much shorter time, according to Chowdhry.
“If a company has a business transformation project going on, and automation is a piece of it, it’s going to spend a third of the [automation] budget just to get that current-state assessment,” Chowdhry says. “And it needs that, because if it doesn’t know what it’s doing now, it can’t improve on it.”
FortressIQ’s software integrates with the video cards in clients’ desktop and laptop computers — the same technology layer that enables screen-sharing — to essentially record video of what happens on screens as workers perform processes.
The software is necessarily complex. For example, it has to understand a range of non-linear actions, since different workers perform process steps in different orders. Sophisticated data mining technology extracts the actual process steps from all such “noise,” facilitating the creation of process maps that are FortressIQ’s main deliverable to clients.
Fortress doesn’t make explicit recommendations for what clients should automate. According to Chowdhry, the mapping itself “can be transformational for a business, because it allows decision-making based not on who’s yelling the loudest but on actual data.”
FortressIQ partners with the major RPA vendors. “We enable you to identify the processes you could automate, which you do in their systems,” the CEO notes.
But deciding to automate a process is just one outcome clients arrive at after receiving the process documentation. They also might choose to (1) standardize some processes by training everyone to perform them a certain way; (2) reconfigure systems to conform to the way employees perform processes; (3) hire consultants to re-engineer processes; or (4) outsource processes.
“If you don’t have the data, you can’t make any of those decisions,” says Chowdhry.