AFGE streamlines IT support with machine learning

With resource constraints and high organizational expectations, delivering effective internal IT support can be a challenge. And with rapidly changing technologies as well as constantly evolving security needs, seamless internal support systems and processes are finally being recognized as fundamental to business success as well as a major contributor to employee happiness.

Thus, it’s perhaps unsurprising that more and more organizations are taking a closer look at how they’re providing best-in-class internal IT support.  The American Federation of Government Employees (AFGE) is one of these organizations.  Its lean team of 15 supports over 350 employees and over 2,500 activists across the U.S.  

Taylor Higley, Director of Information Services, and his team handle everything from company-wide information security and new technology training to “please fix keyboard and mouse” requests.  They not only believe in being proactive in their IT approach, but that data is key for action. They take pride in understanding the needs of their end users and promptly providing key information to enable the best decision, for both their stakeholders and their own team.

As their support ticket volume increased, it became more difficult to classify issues appropriately. Higley began to realize that they were making decisions to invest large amounts of time and energy on their next IT project based on gut feel or general impressions around where problems may lie, rather than being truly data-driven.

About AFGE

The American Federation of Government Employees (AFGE) is the largest federal employee union, representing 670,000 government workers nationwide and overseas. These employees inspect the food we eat and the places we work, care for our nation's veterans, keep the national defense systems prepared for any danger and much, much more.

"The challenge that we fundamentally face is having good data about how we spend our time because our scope of work is so wide."
- Taylor Higley, Director of Information Services at AFGE


Higley and his team were looking for a way to gain insight into the magnitude and impact of the problems stakeholders were facing.  And more importantly, they needed a better way to surface and prioritize common issues, so that the team could dedicate resources appropriately.

By using Wise to automatically classify the type of issue contained within each ticket upon arrival, Higley could then easily query and understand the trends and effect of certain issues on his operations in real time. Furthermore, he didn't need to spend precious time and resources by having his team of experts manually tag and classify tickets as one of the 80+ categories.

In addition, because the predicted category is shown directly within each agent's ticket view, it has frequently served as a quick cue to better prompt agents on how to resolve incoming issues. This has allowed the team to build macros to quickly respond, saying, “Here’s what we think your issue’s about and we’ll still be in touch, but we think this might help." This has helped drive resolutions even when stakeholders submit tickets off-hours, keeping satisfaction sky-high.

"We do have certain people who are better suited to certain issues, so rather than have tickets sit in queue before they get assigned, we have started using Wise Support for routing."
- Taylor Higley, Director of Information Services at AFGE


AFGE also uses Wise Support to route certain tickets to the agents best suited to handle those issues. Taking advantage of the automatic categorization, Higley and his team are able to respond more quickly and accurately as tickets immediately go to the right agent with expertise in that area.

With Wise Support, AFGE has been able to identify and prioritize support issues more efficiently and ensure fast, accurate resolution by their expert team. Always focused on end user satisfaction, the AFGE IT support team has maintained their 98%+ satisfaction rating as they continue to use automation to drive efficiency and process improvements.