AI is changing where customer service decisions are made, moving frontline employees to more responsible driven tasks.
With more support moving into live customer-agent interactions, information from these can reveal the recurring gaps and problems that organizations need to address.
To ensure accuracy is not deprioritized by demand, CX leaders should consider how to connect live guidance with conversation intelligence, so that lessons from individual interactions can inform wider operational improvements.
Jon Lamont, VP of Client AI Automation at CallMiner, told CX Today that giving frontline teams the right information is becoming increasingly important as AI changes the role of service agents.
"As AI takes on more routine work, the value of a frontline representative increasingly comes down to judgement," he said.
"That means giving agents real-time visibility into the customer, the business, and the knowledge they need to make the right decision in the moment."
Give Agents the Context to Handle the Complex Moments
As AI becomes responsible for more routine requests, the role of frontline agents is likely to become concentrated around interactions that require human judgement and ownership.
This leaves humans to deal with customers whose situations are more complicated to be answered through a standard response.
“If you’re using AI in the right way to take away the mundane conversations, it is going to leave your frontline representatives with more complex, unique interactions,” Lamont noted.
If the value is focused on giving an agent the information needed to make an informed decision, this changes what useful AI support looks like during a live conversation.
In fact, CallMiner’s 2026 CX Landscape Report found that almost all organizations (95%) agree that human agents add more value than automation or AI in at least one type of interaction, particularly when complex problem-solving, significant business consequences, customer vulnerability, or empathy are involved.
As AI is increasingly used alongside employees during live conversations, the distinction between assistance and control is becoming more important.
Lamont notes:
"A well-designed real-time AI system should guide agents toward the right questions, relevant policies, and next best actions, without taking control of the conversation."
Furthermore, without context, an agent may spend the opening part of the conversation asking the customer to explain what has already happened, which adds friction.
Giving the representative access to recent customer history can allow them to start from a position that customers expect from a human employee.
“If a customer who’s called five times in the last week, and that context is automatically surfaced when the call begins, the agent can immediately understand the situation, show empathy, and focus on resolution instead of spending valuable time gathering historical context," he emphasized.
The current role of AI is to provide the context that helps the agent exercise better judgement, while leaving the human responsible for how that knowledge is applied.
