As AI takes on more routine contact centre work, the conversations reaching human agents are becoming harder, more emotional and more draining. In a recent Puzzel Talks episode, vulnerability expert Helen Beaumont Manahan makes the case for why that shift demands a very different kind of investment in people.
Ask most contact centre leaders what AI is doing for their teams, and you'll hear about efficiency. Shorter handling times—fewer repetitive queries. Agents freed up to focus on what matters. All of that is true. But there's a side of this story that gets far less attention: what happens to the humans left handling everything AI can't.
Helen Beaumont Manahan has spent her career at the intersection of contact centre quality, customer experience and vulnerability training. As Director of Client Success and Customer Experience at the National Support Network, she works with organisations across financial services, housing, utilities and the third sector to help frontline teams handle the hardest conversations with confidence. Her view on the AI shift is measured but pointed.
The emotional register of frontline work has changed
When AI handles balance enquiries, delivery updates and FAQ-level queries, what's left isn't a lighter workload. It's a concentrated one. Advisors are increasingly fielding customer confusion, distress and, in some cases, genuine crisis - contact after contact, with little space in between.
"Advisors tell us it's relentless," Helen notes. "And they are absolutely feeling it." The industry has long talked about compassion fatigue and empathy fatigue. But as automation accelerates, those pressures are intensifying in ways that many organisations haven't yet accounted for.
The risk, as Helen describes it, is a lag. Investment in AI technology is moving fast. Investment in the human infrastructure that supports the people working alongside it often does not keep pace.
Shorter training times aren't the only good outcome
Puzzel's own research found that close to 90% of CX leaders believe AI will shorten agent training times - and that's largely a positive development. AI tools are already reducing cognitive load, surfacing information at the right moment and removing the need for advisors to hold entire knowledge bases in their heads. That frees up mental bandwidth for the conversations that actually need it.
But Helen pushes back on what happens to the time that's saved. "I really hope that the 30% of training time that's getting freed up is being reinvested into human skills - and not just saved as a cost."
The human skills she's referring to aren't soft add-ons. They're the ones that determine whether a customer in real distress feels heard or processed. How to slow a conversation down when someone is struggling. How to hold silence without filling it. How to notice when what a customer is saying and what they actually need are two different things. These are learnable, buildable skills, but they take dedicated time and practice to develop. They can't be shortcut by a co-pilot.
Where organisations are falling short
Helen identifies three consistent gaps in the organisations she works with.
The first is training for the process rather than people. Advisors know the system steps, but they don't have a framework for what to do when a customer's situation doesn't fit neatly into a category or for how to look after themselves after a difficult contact.
The second is unclear escalation pathways. When a situation exceeds what an advisor can handle, the route forward should be obvious and seamless. Where it isn't, handoffs can feel abrupt and disorienting for the customer — and disempowering for the advisor.
The third, and the one Helen feels most strongly about, is the absence of a debrief culture. "Advisors can finish a genuinely difficult contact and be expected to go straight into the next one. That's not sustainable. It takes the compassion we're asking people to demonstrate and often score them against."
A practical place to start: structured debrief
For contact centre leaders wondering where to begin, Helen's recommendation is straightforward. Introduce a short, structured debrief practice after difficult calls, if one doesn't already exist. It doesn't need to be elaborate or time-consuming. It can run over Teams, Slack or face-to-face, depending on how the team operates.
Done consistently, it achieves several things at once: it normalises the idea that difficult contacts take a toll, builds a culture where challenges are talked about rather than pushed down, and gives advisors a sense that their experience matters - not just the output.
"A team that debriefs regularly is a team that builds collective resilience," Helen says. "And that resilience pays back in retention, in consistency, and in the quality of the next difficult conversation."
Empathy alone isn't enough
One of the clearest points Helen makes is also one of the most underappreciated: empathy, on its own, isn't a strategy. It needs infrastructure to act on. That means clear signposting, defined escalation routes, coaching cultures and debrief practices that give advisors the tools to translate good intentions into confident, consistent action.
As AI continues to raise the emotional stakes of frontline work, the organisations that get this right won't just see it in their customer satisfaction scores. They'll see it in their retention figures too.
Listen to the full episode of Puzzel Talks to hear Helen's perspective on vulnerability as a dynamic rather than a fixed state, and what genuinely good customer outcomes look like beyond a QA scorecard.
