A Conversation with Mike Lytle, CEO of TP USA
In a world racing toward full-on AI everything, the most surprising thing might be this: emotional intelligence is becoming more important, not less.
That was one of the big takeaways from my recent conversation with Mike Lytle, CEO of TP USA (formerly Teleperformance), one of the largest and most quietly influential companies in the customer experience space. With nearly half a million employees operating across 170 markets in 300 languages, they’ve got a clearer view than most of how customer relationships are evolving – and how AI is reshaping the entire field.
Founded in 1978 as a traditional customer care company, TP USA has evolved into something much broader: a global service provider offering everything from interpretation and localization to visa processing, health advocacy, and trust & safety. But what makes them stand out isn’t just their scale – it’s how they’re thinking about AI.
Lytle summed it up simply: “Powered by emotional intelligence, enabled by artificial intelligence.” That’s not just a tagline. It’s a strategic stance – and it reflects a clear belief that while AI can handle tasks, humans still handle relationships.
TP USA isn’t trying to replace agents with AI. They’re using AI to make agents better.
Take TP Interact, for example – a system that uses AI to read emotional cues in real time, analyzing both customer and agent sentiment during a conversation. That data feeds into coaching dashboards, allowing managers to train not just for accuracy or efficiency, but for empathy. It’s EQ at scale, which is a lot harder to achieve than it sounds.
Lytle was also refreshingly practical when it came to AI deployment. His advice? Don’t try to “transform” your entire organization all at once. Start small. Prove the value. Build from there. It’s not a tech problem – it’s an infrastructure and change-management problem. And without clean, governed data, you’re not going anywhere fast.
One of the most compelling parts of our conversation was about agent enablement – the often-overlooked area where AI can actually solve something painful. Most frontline agent turnover happens in the first 90 days, often because they feel underprepared and overwhelmed. TP USA’s answer? AI-driven training simulations and generative AI tools that surface the right answers instantly – no more digging through outdated knowledge bases while a customer waits (and fumes).
Looking further ahead, we talked about agentic AI – systems that don’t just advise or assist, but actually take action across multiple systems. Think: one keystroke instead of ten screens to process a sales order. The time savings are massive. The human value? Even bigger – agents can spend more time building trust, not toggling tabs.
But none of this works without governance. Lytle was clear: the pace of innovation can’t come at the expense of control. Every shiny new AI tool needs to be vetted, validated, and monitored. Because in the real world, hallucinations and bad outputs aren’t just funny – they’re a liability.
In the end, TP USA’s story is a quiet rebuke to the idea that AI will kill off traditional customer experience. If anything, they’re proving the opposite: that smart, ethical, human-AI collaboration can drive better outcomes for customers and better careers for agents.
It’s not hype. It’s happening. And it might just be the most important CX story unfolding right now.
