What’s interesting is that language rarely remains just a CX problem.

Voice is one of the most complex channels in customer interaction. It carries emotion, context and trust. When language gets in the way, the impact is immediate and visible. What is less visible is the ripple effect behind the scenes.

Language shapes hiring strategies, training models, cost structures and service coverage. It influences which markets are prioritised - and which are quietly considered too complex to scale. Over time, organisations adapt to these realities pragmatically. They manage them well. But they are still managing around the constraint.

As enterprises now move toward shared, global operating models supported by AI, that constraint becomes harder to ignore.

At VOIS, we sit close to both customer experience and large scale service delivery. That perspective made one thing clear to me early on: continuing to ask “how do we add more languages?” would only take us so far. So we changed the question.

What would genuinely change if language stopped determining where and how a service could be delivered?

That shift in thinking led us to focus on building VOIS as a language agnostic service provider - not as a theory, but as a practical enabler of scale. This is where we launched our Real Time Language Translation (RTLT) programme: to create a capability fit for Vodafone - one that is commercially viable, operationally scalable across geographies, and adaptable to enterprise environments without heavy integration overhead.

The intent was simple: remove language as a limiting factor in how global services are designed and delivered.

Voice leaves very little room for error. In live conversations, even small delays feel significant. Any sense of artificial mediation breaks trust quickly. That’s why my view has always been that AI in voice environments must be almost invisible. It needs to work in real time, preserve the natural flow of conversation, and stay firmly in service of the people speaking - not the other way around. Achieving that balance hasn’t been easy, but it has been essential. When the technology feels natural, the focus returns to understanding and intent - not translation.

Reducing language friction in live conversations has implications that go far beyond CX.
It allows services to scale without duplicating capability market by market. It widens access to talent by prioritising skill and empathy over native language alone. It makes smaller or more fragmented markets viable in ways they often aren’t today.

For VOIS, this aligns closely with how we think about shared intelligence: designing capabilities once and scaling them across customers, sectors and geographies. Language‑agnostic service models don’t remove complexity entirely - but they stop language from being the deciding factor in every growth conversation.

AI is driving many incremental improvements across enterprises. But some opportunities are more disruptive in nature. Addressing language at a structural level is one of them.
My belief is that organisations willing to tackle this thoughtfully - without compromising the human experience will move faster, scale with greater confidence, and operate more flexibly as work becomes increasingly agentic.

At VOIS, we’re proud to be shaping that conversation - not by talking about AI in abstract terms, but by redesigning how global interactions happen.

One agent. Any language. Zero friction.