Meet ServiceNow Agent Intelligence: Machine Learning for Everyday Work
UPDATE: Agent Intelligence is currently available on the Now Platform ® .
CIOs are finally cutting through the machine-learning hype. According to “ The Global CIO Point of View ,” issued today, nearly 9 out of 10 CIOs already use machine-learning technology or have plans to adopt it. Investments will double in the next three years as leading organizations turn to advancements in machine learning to drive faster, more accurate decisions that fuel digital transformation.
A shortage of talent has hindered adoption. Designing and rolling out practical applications of machine learning has required highly trained data scientists. Recent analysis shows that the demand for data scientists far outstrips the supply. In fact, only 27% of CIOs have hired enough talent to support their machine learning plans.
ServiceNow is introducing a breakthrough approach to make machine learning more accessible to organizations –one that does not rely on an army of data scientists. We have infused machine-learning capabilities directly into the Now ® Platform, making it accessible to all our applications.
Agent Intelligence, our first machine-learning solution, will be available in the next release, Kingston.
Agent Intelligence will automate the categorization, prioritization, and assignment of work to reduce resolution times, minimize human error, and improve customer satisfaction.
Agent Intelligence will initially be applied to improving the quality of IT and customer-service processes . Most companies have an IT service desk as well as a customer-support center. These functions ensure that customers and employees receive help in a timely manner—whether it is a request for a new computer, a password reset, or the status of a repair. Manually triaging thousands of incidents or cases, categorizing tasks, and routing work can consume people resources, introduce human error, and slow resolution.
In fact, a new report from Accenture found that front-line customer support functions spend up to 12% of their time categorizing, prioritizing, and assigning tickets. And 43% of IT service desk respondents are weighed down by having to choose from 100+ assignment groups—with nearly a quarter needing to choose from 300+ assignment groups.
Agent Intelligence for IT Service Management and Agent Intelligence for Customer Service Management were designed to address these very challenges. Specifically, these solutions:
- Categorize and Prioritize Work Faster – Agent Intelligence machine-learning capabilities accurately categorize requests nearly instantly. This eliminates the bottleneck created by agents having to manually triage, categorize, and route requests (incidents/ cases). A requester no longer has to scroll through multiple lists and rely on organizational knowledge or guesswork to choose the right category or assignment group.
- Decrease Time to Resolution and Error Rates – Automatic routing means tasks and incidents get the attention they deserve from the right individual, increasing overall customer satisfaction. Agent Intelligence quickly classifies and routes requests with fewer errors, increasing agent productivity.
Customers that piloted Agent Intelligence have reported impressive results . The pilot program revealed that, on average, customers can expect to save 8% of their service desk’s time through improved categorization, prioritization, and assignment of incidents.
The feedback from the pilot program has been extremely positive:
“Of the incidents categorized by ServiceNow Agent Intelligence, we can expect a 97% to 99% incident-assignment accuracy.” -Senior IT manager at a major retailer
“Our aim is to reduce the number of low-value tasks and focus our service-desk agents on higher value tasks in IT that drive value for the business, as opposed to logging tickets.” -Customer support technology and operations manager at a multi-national energy company
ServiceNow’s own Now on Now program also put Agent Intelligence through its paces. The customer service team handles more than 13,000 incidents a month. In our tests, Agent Intelligence has been 95% as accurate as humans are in categorizing the incidents. Thus, the team has saved more than 315 hours in just two months from productivity improvements and has seen a rise in customer satisfaction.
Agent Intelligence is a great example of the practical application of machine learning to getting everyday work done faster and smarter, without having to depend on scarcely available, highly skilled data scientists.
Learn more about Agent Intelligence here .
