The AI-powered virtual agent integrated into Jira Service Management enabled Atlassian to scale its IT service desk faster. Learn how it could benefit yours.
Atlassian IT’s ticketing system is built on the back of Jira Service Management, a powerful tool that enables service desks to be set up quickly and adjust to any scale. With an intuitive interface on Jira Service Management, service desk agents can create threads to manage requests, add comments, and interact with customers filing requests in an easier way. The product has become the core of service desks for many Atlassian customers.
Atlassian’s acquisition of Percept.AI in 2022 strengthened Jira Service Management’s conversational support offerings. Percept’s virtual agent technology already focused on the conversational experience, but integrating it with Jira Service Management creates an AI-powered layer to enable better conversational experiences in portals and chats. For companies in the midst of a transformation and scale, the Jira Service Management virtual agent feature is another tool available to them.
Learn how Atlassian IT leveraged the virtual agent to save hundreds of hours of work for its service desk team, and use their story to find opportunities to streamline your service desk with Jira Service Management virtual agent.
Upleveling the IT service desk
Good IT service should be second to none in the distributed work world. That’s why Atlassian revamped its service desk at the start of the pandemic to streamline and improve the quality of service for remote employees.
That revamp was just the first step, but it worked like a charm, moving requests from disparate channels into one Slack channel that was tied to a Jira customer portal.
But the team wasn’t staffed to offer real-time support to employees as Atlassian rapidly scaled to more than 10,000 global workers in the past two years. We needed to find a cost-effective, scaleable solution to support the growth.
Embracing an innovative spirit, Atlassian created an AI-powered virtual agent that instantly gathers information and resolves simple support requests 24 hours, 7 days a week, 365 days a year – all without any involvement from human support staff.
The technology powering this solution is a result of a collaboration between two of Atlassian’s acquisitions of Percept.AI and Halp. As essential parts of Jira Service Management, the technology enables an artificial intelligence-powered virtual agent designed to help companies create better and faster responses to employee service desk queries at scale.
Automating service and help desk systems
Company growth should be at the forefront of planning for IT support. It was the reasoning behind Atlassian IT’s consolidation of service desks into a single Slack channel for all information technology-related requests.
“Imagine that you have a service desk that handles 10,000 support tickets a month,” says Joe Flowers, technical program manager for the team that manages Atlassian IT’s service desk. “Everything from hardware requests, network issues, access issues, configuration changes, and more. On top of that, over 10 percent of them are resolved by the virtual agent in Slack without requiring human intervention.”
While the feature is in its infancy, it has already created a meaningful impact across Atlassian IT’s operations. The time saved in the first month was an incredible 500 hours – 20 days – in how long a service desk team would spend finding information and providing repetitive answers to tickets.
By the numbers, the virtual agent provided assistance with over 50% of all incoming requests, either resolving the issue or gathering more context for a human support agent. As part of that, the bot fully resolved over 10% of all requests, translating to 500 fewer tickets that needed to be answered by a real person.
By integrating an AI layer into the process to answer the repetitive but necessary questions that come every day, service desk agents will realize incredible benefits.
Building an IT service desk to scale
The Jira Service Management virtual agent relies on pre-configured, data-driven intents to automatically decipher employee requests and direct them to specific knowledge base articles deemed relevant by the team.
“With this virtual agent, we’re able to assess the customer’s request and, if the request is about a topic in our knowledge base, we’ll just automatically serve that up directly in Slack,” Joe says.
Employees that interacted with the bot found it incredibly useful, rating it an average of 4.5 out of 5 on customer satisfaction scores. Even more impressive was the time saved.
It means less time spent answering questions about VPN issues or logging onto Zoom and more time focused on more complex projects that require more nuance and strategy. For complex tickets where an article won’t suffice, the bot can recognize the need for a human to intervene and seamlessly hand it off to a real service desk agent monitoring a queue for complicated asks.
In theory, the bot could serve the needs of an entire organization looking to answer level-1 – meaning the repeatable and documentable – support questions.
“If you think about how many repetitive questions we get every week, it’s a lot. This AI-powered virtual agent is meant to take on those important but monotonous questions and answer them in less than a minute,” Joe says.
From homegrown answers to enterprise-level solutions
Why build this solution from within when there are various virtual agent solutions available on the market? For Atlassian, the answer was that solving a problem using Atlassian products means other companies and teams can reap the benefits.
The organization triages an average of 30,000 tickets each month, allowing a proper stress test for products at an enterprise level.
The program’s success in solving IT service desk questions spawned interest in integrating the virtual agent into other aspects of the company. Atlassian’s human resource team adopted the same virtual agent technology to handle questions about stipends, visa statuses, filing expenses, and more.
“We’re not just doing it for the sake of doing it. Atlassian is big enough that we are literally the target demographic,” Joe says. “So why wouldn’t we use our solutions to solve the same problems that Atlassian’s customers face?”
Interested in implementing a virtual agent to your service desk? Join the waitlist for the beta launch.