Thousands of students for George Clooney at the Cuneo Sports Arena

Every year Fondazione CRC brings an internationally known figure to Cuneo for "Dialoghi sul Talento": before George Clooney, the series hosted Kerry Kennedy, Paolo Nespoli, Andrea Bocelli, Pep Guardiola, Steffi Graf and Andre Agassi. The 2026 edition, devoted to human rights and produced with the Clooney Foundation for Justice and Collisioni, moved the event into a sports arena to make room for thousands of high school students from across the province.

The Challenge

The constraint was not linguistic complexity. A single language pair — the guest speaks English, the audience is Italian — but in front of a crowd the size of a sports event, in a venue never designed for conferences.

Receivers do not scale. Handing out thousands of receivers to underage students means managing distribution, collection, deposits, hygiene and loss, within entry and exit times a school morning simply does not allow. Hardware alone was not a viable route.

An arena network is not a congress centre network. Thousands of smartphones switched on inside the same volume is a saturation scenario, on Wi-Fi as much as on cell towers. BYOD distribution with no alternative would have concentrated the entire risk on a single point.

Then there was the audience itself. A school crowd listens unevenly: very different levels of English, a reverberant room, content covering civil rights and international affairs. The audio channel alone does not reach everyone. And accessibility was a requirement, not an option: live captioning did not arrive as an upsell, it is the standard American productions work to, where captioning is taken for granted at any public event.

Finally, the voice had to stay human. A talk on rights and politics in front of students, with irony, jokes and implicit references, is exactly the territory where human interpretation cannot be replaced. All of it on a single date, with a guest available on a single date: no window for a second attempt.

The Solution

Two distinct layers, not to be confused: the spoken translation is human, the on-screen text is AI. Underneath, distribution across three parallel channels, sized so that none of them would be a single point of failure.

Simultaneous interpretation, human voice

The translation students heard in their headphones was delivered by professional conference interpreters. No synthetic voice on the audio channel: in a talk built on irony, understatement and political references, quality depends on a reading of context only a human interpreter provides. The criterion is the usual one — AI where content is frontal and technical, interpreters where there is debate, irony or sensitive material — and this event fell entirely on the second side.

Receivers: 400 units where listening had to be guaranteed

The receiver fleet was deliberately kept low and concentrated where listening had to be independent of the network: chaperones, guests, teachers, assigned sections. It is not the main distribution channel, it is the one that works regardless.

Converso® Web App: BYOD for up to 3,000 users

Most of the audience listened to the translation on their own smartphone, with their own headphones, by scanning a QR code. No receivers to hand out and collect, no queues at the entrances, no battery and sanitising logistics across thousands of devices. The 3,000-user sizing was chosen on the venue's actual capacity, not on the expected turnout.

Phone dial-in backup

Alongside the Web App, traditional phone audio access was enabled. If the local data channel degrades, listening remains available over a voice call. It is an unglamorous redundancy, and a very useful one in a packed sports venue.

AI live captions: accessibility first

The on-screen text was generated by the Converso® AI Engine with human governance: a technician monitors stream quality and acts on the causes of degradation.

The reason it was there is accessibility. Anyone with hearing difficulties follows the event without having to declare it to anyone and without a dedicated device — in an audience of thousands of underage students, that is the difference between a service being available and a service actually being used. It moves accessibility from the client's goodwill to the technical specification, where it can be designed and verified. Everything else is a benefit that comes after, not before: the text gives a second foothold to those who follow English only partially, and keeps proper names, acronyms and geopolitical references in focus.

On-site crew: 2 technicians, 2 days

A full day dedicated to setup and rehearsals on 7 April, and the event day itself on 8 April. In a sports venue the variable is not the rack: it is the placement of audio pickup points, performance in a reverberant room, and interfacing with the AV control already in place.

The Results

The event ran to schedule on 8 April 2026, with thousands of students in the stands. Clooney opened by greeting the audience in Italian and then spoke in English on rights, with headphone translation delivered by human interpreters throughout.

Listening was guaranteed across three independent channels — receiver, smartphone, phone dial-in — plus the text channel of the captions. No receivers were handed out to the student audience: thousands of teenagers entered and left with no distribution, collection, deposits or queues at the gates. And accessibility was covered with no self-declaration and no dedicated hardware: those who needed the text already had it on screen.

For audiences above a thousand people, the right question is not how many receivers are needed, but how many listening positions must keep working if the network fails. The cost of phone-based redundancy is marginal compared to the cost of a single-date event that cannot be run again.

Cover photo: San Rocco Castagnaretta Sports Arena, Cuneo — by Luigi.tuby, via Wikimedia Commons, licensed CC BY-SA 3.0.