Blog Articles

A meeting can look successful in the moment and still fail the people in it. Everyone nods, the agenda gets through, and the call ends on time. Then the follow-up emails arrive, and it becomes clear that half the room understood something different from what was said. In multilingual settings, this gap is rarely caused by a single dramatic failure. It tends to come from a handful of small, hidden weaknesses in the speech translation platform and the way it is set up.

 A product launch, a leadership announcement, a crisis response — whatever brings journalists and cameras into the room, a press conference is one of the most high-stakes communication moments your organisation will manage. Every word matters. Every misquote matters. And increasingly, every language matters too. 

Quoting audio translation can feel like comparing apples to oranges. One provider prices by the minute, another by the language pair, a third bundles everything into a flat event fee. For an events manager trying to budget a multilingual conference or hybrid town hall, the lack of a standard pricing model makes it hard to know whether a quote is reasonable before you've gathered several.

Imagine you’re at a multilingual event where interpretation was available and everything seemed inclusive, but you still missed a sentence and could not go back or follow everything was said. Sounds familiar? Perhaps the speaker moved too quickly, the audio dipped for a moment, or you were taking notes when an important point was made. At live events, even strong interpretation and good production do not always prevent information from being missed.

Think about the last large event you attended or organised. How many of the people in that room — or joining virtually from around the world — were expected to simply follow along in a language that wasn't their own? 

Speech to speech translation is the process of converting spoken words in one language into spoken words in another language. Instead of asking participants to read translated text, the system delivers translated audio that helps people follow a meeting, event or presentation in their preferred language. 

Every word matters in football. A manager’s explanation after a defeat, a captain’s message before a final, or a new signing’s first words to the press — these moments are broadcast across the globe, shaping headlines and club reputations.

But football isn’t spoken in just one language. From La Liga to the Premier League, from Serie A to the Champions League, clubs welcome players, journalists and broadcasters from every corner of the world. And UEFA’s regulations make sure those voices are included: press conferences must be accessible through simultaneous interpretation.

The race to build AI-powered multilingual communication took another major turn this month as Cisco finalised its acquisition of EzDubs — the Y Combinator-backed real-time translation startup known for its voice-preserving technology. Cisco had initially announced its intent to acquire the company in November, and EzDubs will now be folded into Cisco’s Collaboration portfolio, including Webex, signalling an acceleration of Cisco’s ambition to make multilingual communication a native feature of its ecosystem.

The multilingual communication landscape is shifting quickly. Cisco’s recent intent to acquire EzDubs, a startup known for speech-to-speech translation that preserves tone and voice characteristics, is the latest signal that major platforms are accelerating their push into AI-driven language services. If the acquisition is finalised, Webex will eventually benefit from more natural, voice-preserving translation technology.