YOUNGSTOWN, Ohio – Information on customers. Information on inventory. Information on cash flow. Information on marketing campaigns. Information on vendors. Information on paperwork. The amount of data a business can catalogue is nearly limitless. With the right tools, that data can help to improve efficiency and productivity. And increasingly, the tools to do so incorporate artificial intelligence and machine learning.
“It’s becoming cheaper to build the model, both from the computing perspective and from the time perspective, so the barrier isn’t there any more. The real trick is how you apply it,” says Quentin Fisher, founder of HC Analytics.
The company, based in Cleveland, is a Youngstown Business Incubator portfolio company that developed a system that uses artificial intelligence to help medical offices predict when patients will miss appointments, allowing the practice to keep its schedule as full as possible. In some cases, Fisher says, offices have been able to increase their capacity 20%.
While HC Analytics’ platform, AIdan, is tailored to the medical field, the basics of how it operates are much the same as other platforms that use artificial intelligence and machine learning. After a model is developed, the user plugs in the data he has and the program picks up on trends and patterns, allowing it to model future results. The more data that the program is given, the more accurately it can predict.
Another application of AI is in enterprise resource planning software, which collects, stores and analyzes data and can automate some internal processes.
“In times like these, with so much uncertainty, it’s useful to look at payment history. The system analyzes payment history of customers so businesses can better predict their own cash flow. It can predict, based on the past, what you can expect,” says Pat Cioffi, vice president of NEO3.
The company in Canfield is a provider of the NetSuite ERP platform. Through its available suite of tools, the software can be used in everything from sales and marketing to inventory and shipping. When it comes to sales, Cioffi offers an example: A system using AI could examine what online shoppers are looking at and what similar users buy to show the shoppers items that they’re most likely to buy.
“It can put those items right in front of them because it’s what they’ve looked for before or because similar people are interested in it. All of it uses learning that these systems do,” he says. “It works the same as marketing. It can optimize marketing campaigns to show you what worked in the past: what they clicked on, what webinars they watched, where leads are coming from. It takes data you’re already using and combining it with technology to create more profit.”
Artificial intelligence can also be used to smooth out internal processes. At accounting firm Packer Thomas, director of information technology consulting Jeff Sheets points to the software the company sells to clients – Accumatica – and its implementation of machine learning on the platform’s help pages. If a user searches for help on, say, purchase orders and closes the first two results quickly but spends 10 minutes reading the third, the system remembers and pushes that result to the top for the next search.
And rather than sift through spreadsheets, Accumatica also features integration with Alexa and Siri, allowing users to use the devices to quickly access data.
“You can ask it questions about a customer’s balance or inventory and it’ll tell you. If it’s too much information – if you ask for the inventory of a specific product by each of your locations – that may be too much for Siri or Alexa to convey. But the software knows that and can display it on your screen,” Sheets says.
Implementing Accumatica is as simple as migrating the data over using an Excel spreadsheet or CSV file, he says.
“You’re not changing how you do everything by switching from one ERP system to another. How you run your business is still how you’re running your business,” Sheets says. “What’s happening is implementing the software that can best serve you. The key to these newer solutions is better functionality that you didn’t have in your existing system.”
Over the past three years, PNC has implemented artificial intelligence or machine learning into more than 100 of its processes, says Dan Pavlick, vice president of strategic services and enterprise architecture at PNC Financial Services Group. One of the most straightforward and easiest to understand, he says, is its use in the bank’s online bill pay system. If a payment has a return because of something like an incorrect address or account number, the system can match up the correct data and continue to process the payment.
“It used to take five to seven days to resolve that issue. Now, because there’s so much data acquired – making sure we’re solving for the right customer, where the data is located, where the vendor changed something – we can use those processes to take that time down to less than a day,” he says.
For businesses, AI in the processing system can route payments through the shortest path to their destination, which can reduce the cost of sending them – especially if they’re international – because each stop along the way adds fees.
“There are multiple paths. So if you model out to suggest back the best way from Point A to Point B, it can take into consideration the timeliness of settlement and the costs associated with it,” he says. “We can find those savings and pass them along.”
AI is also used to detect fraud and onboarding new clients, both consumers and businesses, Pavlick says. For each use – whether at a small machine shop or a national bank – the most important thing to have is data, Pavlick says.
“If you want to do something, you need the data to train the model. When you look for historical data [to do that], you may look back and find some of the data that allows for the model; but you don’t have all of it,” he says. “It might be something that at one point in time you viewed as disposable or, in a more analog world, was less important.”
And for any size business, HC Analytics’ Fisher says, adding AI and machine learning tools can be a benefit. Platforms like NetSuite or Accumatica can be customized to every business, either internally or through industry-specific additions, and be used in just about whatever way the business sees fit. With the capabilities becoming greater and the costs dropping, the tools are needed by businesses that want to stay on top of the competition.
“It’s not that expensive and it can have an incredibly high business return if you have a real problem. … They’re quite economical to build and there are organizations like us that will help you do it. Once they’re up, there really isn’t much maintenance at all,” he says. “I can do more with three people today than I could have done with 30 people three years ago. It’s moving that fast.”
Pictured: An application of artificial intelligence is enterprise resource planning software.