Tablet-based point of sale terminals may look sleek and modern, but under the hood they must wrestle with many of the same challenges that older hardware faces.
To keep things running smoothly, machine learning — a form of artificial intelligence — can play a major role, said Bobby Marhamat, chief revenue officer of Revel Systems, which builds iPad point of sale management technology for retailers.
"With tablet technology a lot of things can go wrong. The receipt paper can run out, the devices can lose internet connection, or anything that gets in the way of focusing directly on the customer," Marhamat said. "It's useful to fix these things automatically, or anticipate them."
Revel just launched RevelGuard, a hardware device that monitors the health of a payment network and builds alerts for point of sale terminals that use Revel's technology. Its technology "learns" to spot specific troubles with a merchant's payment system, automatically creating a maintenance ticket before the merchant notices there's something wrong.
The system also handles more mundane tasks such as ensuring software updates are property deployed across a retailer's network, and sends an alert if there is a mismatch. The technology learns to spot these issues based on prior maintenance requests.
"There's a lot of things that can happen," Marhamat said. "A merchant can get an alert from a virtual agent telling the merchant he or she is about to run out of receipt paper, or to check the cable or Internet network."
By using machine learning, RevelGuard is trying to make maintenance more intuitive by giving it a memory. For example, one of the early adopters, Burger21, used Revel's technology to discover a WiFi network glitch was responsible for the frequent "timing out" of its point of sale terminals.
The concept of machine learning has existed for years, but improvements in the underlying technology and a need for fast action to accommodate faster payment processing and m-commerce are giving the technology a life in the financial services and payment industries.
AI is seen as improving compliance tasks such as anti-money laundering and 'know your customer' requirements. Among payment companies, PayPal is using an internally-build AI engine to help separate false positives from truly suspicious transactions or accounts. Payment startups are also using AI to broadly reduce friction for user experience, speed to market and product development.
But despite the increasing popularity of the technology, there is still a learning curve, particularly among payment executives who don't understand that innovation, though are confident it can help wring inefficiencies out of payment processing and risk management.
Monitoring systems have become a staple of the IT infrastructure today and the more advanced solution sare using machine learning to monitor computer log and performance files, networks and peripherals to not only recognize problems but to take corrective action, said Tim Sloane, vice president of payments innovation and the director of the emerging technologies advisory service at Mercator.
"These tools will monitor the point of sale and the payment related transactions with minimal changes as long as the point of sale keeps a log file, with out personal information or card data," Sloane said. "That said, Revel sells to small merchants that are unlikely to have sophisticated hardware and network monitors so offering such a solution is clever and may help generate additional revenue from its customers for the service."