When a global business services company grows fast, internal support, like HR, IT and payroll — has to scale with it. Otherwise, employees waste time waiting for answers, and support teams get bogged down with repetitive manual work. To solve these problems, the company built a chatbot to automate processes.
The challenge
Chatbots hit their limitations
Adoption of the legacy chatbot sat at 5%. The tool was limited to rigid, keyword-matching logic only. Without the ability to understand context, it failed the moment an employee asked something slightly outside the script — which was most of the time. Employees quickly deemed it ineffective.
When the bot couldn’t answer a question, it generated a support ticket — without having asked follow-up questions. Tickets arrived half-empty, and human agents spent their time chasing missing information across multiple systems. Technology that was supposed to save time was eating it instead.
The solution
Agentic AI opens new possibilities
We replaced the rules-based bot with an agentic AI platform that’s capable of reason through problems. Built on Decagon technology, the new agent learned from more than 2,000 knowledge articles across 17 support functions, which ensured enough depth to resolve a much broader range of employee inquiries.
With advanced reasoning capabilities, AI agents can ask questions and attempt to resolve issues on their own before creating a ticket. The technology knows exactly when to pull in a human.
Automation also solved the adoption issue. The tool connects directly to the manual ticketing portal, so anyone who tries to file a ticket gets routed through the AI agent first.
Results
Saving time and delivering proper support
With agentic AI handling routine inquiries from start to finish, the company saw a 57% reduction in tickets, allowing teammates to focus on high-value, complex work.
Today, the system provides seamless, real-time support across all 17 support departments.
What our experts say
How Agentic AI Changes Frontline Work
There’s no doubt that agentic AI drives significant efficiency gains — so much so, the technology is changing who handles support contracts and how they’re being written. But there is an often overlooked tradeoff that companies need to consider.
When AI handles routine work — the repetitive, predictable and easily scripted — human agents get a harder job. Each call reaches them because the AI couldn’t manage it. These new challenges can be both energizing and exhausting.
“For those agents who really like to solve problems and handle more complex work, it’s better work,” says Chris DeLambo, Division Vice President of Agentic AI Solutions at TaskUs, in an interview with The Financial Brand. “It draws on more of your knowledge, it’s more exciting, more interesting — it’s a better job in many ways.”
Better, but harder. In the old model, simple calls acted as natural cycles of relief. Agentic AI eliminates most of that breathing space, leaving a concentrated stream of high-stakes conversations that demand more patience, judgment and empathy. The frontline needs a new kind of support too.
Forward-thinking organizations recognize that mental health resources are now an operational necessity and are responding by providing science-backed wellness programs.
For example, TaskUs employs more than 200 licensed clinicians to support its frontline staff. “These professionals are working with our agents and our frontline staff to help them with the mental health part of their work,” Chris explains. “We see that as a key component as we get deeper into AI and deeper into these complex, emotionally charged conversations.”
Read the full article for more insight on how agentic AI impacts the frontline.
