ServiceNow Kingston troubleshoots IT tasks with AI, machine learning

ServiceNow delivers AI and machine learning capabilities to its core platform to eliminate time-consuming manual help desk tasks.

ServiceNow has seasoned its core platform with AI and machine learning technologies in the hope that it will automatically route technical problems to IT professionals best equipped to handle them.

The Agent Intelligence contained in the latest release, code named ServiceNow Kingston, is a supervised machine learning application designed to reduce the manual processing of help desk requests, and thus reduce IT help desk backlogs that face many shops today. The debut of the DxContinuum technology ServiceNow acquired in 2017, identifies and examines benchmarks within the IT infrastructure and predicts trends and events based on data collected through operational intelligence tools.

The development of ServiceNow Kingston's AI and machine learning technologies was guided by research the company conducted among users. Above all else, the majority of them want assistance with everyday IT tasks, according to Allan Leinwand, ServiceNow's CTO.

"We learned many users don't have the in-house talent, such as data scientists or others fully indoctrinated in AI, to apply it," he said. "They don't want to have to learn machine learning in Amazon or Microsoft's public clouds. They want to use it in a way that's practical for them."

However, to ensure the new technologies deliver accurate results, users must have at least 50,000 incidents or pieces of data, which means the product is best suited for midrange and larger IT shops.

Allan Leinwand, ServiceNow CTOAllan Leinwand

"We found that figure to be the tipping point where they can get a high degree of accuracy, and get the most out of the system," Leinwand said.

To illustrate how incident categorization, prioritization and routing capabilities work, Leinwand described a manufacturer of aircraft engines that develops a customized application that contains thousands of pieces of data to track and measure various performance metrics of engines.

To ensure the new AI and machine learning technologies deliver accurate results, users must have at least 50,000 incidents or pieces of data, which means the product is better suited for midrange and larger IT shops.

"Through the ServiceNow platform, users can look at the table of information called aircraft engine performance, look at the column listing the number of hours in flight and generate back a prediction about the number of hours of flight time left before it needs to go into service," Leinwand said.

Another addition to ServiceNow Kingston is Flow Designer, drag-and-drop software that lets non-programmers assemble process flows for projects. The software works in concert with the company's existing Workflow editor and across ServiceNow's product portfolio, as well as users' third-party applications. Another addition to ServiceNow Kingston is the Integration Hub, which orchestrates the interaction between Flow Designer and a range of third-party products.

Leinwand said users "can put together a workflow on our platform that will send a notification out to Slack or a group within Microsoft Teams, and the Integration Hub will integrate a workflow with those of third parties."

Through the testing process, ServiceNow discovered that the new technologies are applicable to more markets than originally estimated. For instance, consumer electronics companies with defective products in the field must route problems to appropriate technicians, and then notify potentially hundreds of thousands of users about how they should handle a recall or delivery of in-the-field fixes.

Dig Deeper on Systems automation and orchestration

Software Quality
App Architecture
Cloud Computing
Data Center