Over the last 12 months, we have seen a surge in investment in Artificial Intelligence (AI) enabled customer self-service technologies as brands have put in place tools that have helped deflect calls away from their support teams and allow customers to self serve.

However, despite these investments, we have also seen how the phone is still an important and vital channel for many organizations regarding customer service. According to Salesforce data, daily call volume reached an all-time high last year, up 24% compared to 2019 levels. Meanwhile, Accenture found that 58% of customers prefer to speak to a support agent if they need to solve an urgent or complex issue, particularly during times of crisis.

Now, consider one of those calls.

When a customer gets through to an agent, they are not thinking about how many calls they have already answered that day, what those calls have been like and how it may have impacted them. The customer, in the moment, is only thinking about solving their particular problem.

That’s all very well, you might say.

But, in the face of consistently high call volumes and the strains of working remotely for an extended period, reports are now starting to emerge that many contact center agents are beginning to experience a similar phenomenon to what many nurses and doctors often go through: compassion fatigue. This is the situation where, due to consistently high workloads, they become emotionally exhausted, on the verge of getting burnt out and become unable to deliver a high level of service.

That, in turn, feeds directly through to the service and experience that the patient or customer receives.

However, Dr Skyler Place, Chief Behavioural Science Officer at Cogito, believes that compassion fatigue is avoidable, and organizations should be using AI to enable and support their agents whilst on a call and, at the same time, manage their well-being and performance.

He believes that there are three areas that organizations are under utilizing AI when trying to improve their customer experience (CX).

The first is that brands should be leveraging AI technology to provide real-time feedback whilst an agent is on a call to support and empower them in the moment.

Secondly, given that many support teams are still working remotely, AI technology can replace the tradition of “walking the floor” and help supervisors understand how their teams are doing and what sort of coaching and support they need from call to call.

Thirdly, when you combine that data with customer outcome data and apply AI technology, you can identify insights that, as Place puts it, “can help you improve your business processes, your business outcomes and drive macro strategies beyond the call and beyond the call center.”

The potential of a system that provides both in-call real-time support for agents but also intelligently understands call demand, an agents experience and in-shift call profiles such that it can optimize call matching to help achieve positive customer and employee outcomes is nothing but a good thing.

Compassion fatigue is real, and organizations need to be managing their agent’s performance and well-being if they are to achieve excellent phone-based customer service.

This post was originally published on Forbes.com.