For the leader of a company with strong positions in cloud computing and machine learning, Pegasystems CEO Alan Trefler is unusually blunt about both.
On cloud computing, he warns, “There’s an unprecedented amount of hype and bull and exaggeration that’ going on in the market”.
Meanwhile on machine learning, Trefler tells Which-50, “We have a triumph of hype and nonsense over actual good thinking”.
- Take our Which-50 reader survey and go into the draw for a chance to win a pair of Sony H.ear Bluetooth headphones or 1 of 5 Google Daydream VR headsets
We caught up with Trefler last week during a visit to Sydney, where he discussed the failure of boards and executive leadership to heed some of the lessons of the past.
He also outlined Pegasystems’ latest initiatives in the robotic automation and machine learning space, and offered a quick overview of what to expect from this year’s PegaWorld, to be held in Las Vegas in June.
Un-shroud the Cloud
While stressing his company’s belief in the benefits of cloud computing, Trefler said “Where I think people have been deceived is this idea that anything that’s a cloud is going to be magically innovative, as if the short term is all that matters. So they say ‘We’ve got to get something up in 90 days’. Sure, but they’re not actually thinking about what that is going to look like in two or three years’ time.”
Companies are not only ignoring the problems of information silos, they are reinforcing them, he said.
Trefler believes there is a real danger that organisations have forgotten some hard lessons from 20 years ago, when they were locked in with certain providers. “If you lock into a Salesforce or an Oracle or a NetSuite, they get all the benefits from economies of scale in the cloud. And all you get is lack of choice.”
A better approach, he insisted, is to provide customers with the tools to make choices about public clouds and internal clouds. “That ensures the customers will always have ownership of their business and their data, and will always be able to take advantage of economies of scale that may occur in the cloud,” he said.
“Don’t get me wrong, we think customers should take advantage of the cloud. But I don’t understand the short-sightedness of board members who eliminate choice from their equation.”
His views on artificial intelligence and machine learning struck a similar tone. Pegasystems is making important investments in these areas, and sees significant benefits for companies now and over the longer term. However, he suggests there is a danger than in the hype: too much emphasis is placed on the notion that the machine will solve all the problems.
“If you want to understand how we believe machine learning and machines should work with people, look at the writings of the former world chess champion Gary Kasparov, on this concept of advanced chess — where you take a good player and team them with a couple of good computers.
“In those circumstances, a person can actually out-play the world’s best players — or the world’s best computers — because you’re bringing human judgment together with what the machine can do,” he said.
Last week, Pegasystems announced new artificial intelligence and robotic automation capabilities that flow from its acquisition of OpenSpan a year ago for $US52 million. These initiatives will bring its PegaWorkforce Intelligence offering under its Customer Service and Sales Automation offerings.
“We provide the ability to have a robot on the desktop that’s simply observing and drawing conclusions about the work that people are doing. You can see the applications they are using, how much time they spend in them, whether they’re cutting and pasting between those applications or re-entering information that might be visible someplace else.”
All the while, Pegasystems’ robots are learning and diagnosing real activity and opportunities. “We do this on the cloud, so that we can bring together the information — not just from one or two users, but from hundreds of thousands of users in operation.”
Trefler describes three components to the approach.
The first is the robotic observation that monitors what is happening. “We’re dealing with the reality, not what people want to think is the reality. And that’s independent of any other Pegasystems application. We watch every keystroke, and every movement that happens on those desktops.”
The next step is the opportunity identification. “Having brought these observations together on the cloud, we can aggregate — by types of work they do, for instance — and we do that without needing personal identifiable information.”
The robots can observe patterns and determine what they might be able to do automatically. “The machine can learn elements and say ‘Hey, I can do this for you’.”
Finally, all these elements link back to the traditional Pegasystems case management and process automation tools, which offer a control structure, “… as opposed to having people flailing around all these systems. We synthesise all the data to determine the opportunities and then we also have robotics which can then take those opportunities and operationalise improvements”.
We also asked Trefler what his customers should expect at this year’s PegaWorld.
“We’re doing a lot of work to improve the agility with which customers can deploy our technology. We want to make it easy for them to manage all the processes and elements of customer engagement they will need, but still do this on a system that’s capable of growing to support mission-critical applications.”
He said a theme was the need to keep pushing the boundaries in areas like artificial intelligence, and building on the machine learning pieces that were announced last week.
“Those are the types of things that you’re going to see us highlighting. It is about bringing agility to businesses, but giving business people tremendously more control over the way their technology is performing.”