Live translation for talks

Every listener reads the talk in their own language.

The speaker talks. LinguaLive writes each sentence down as it is spoken and translates it on every phone in the room. No app, no headset.

Where it fits

Any room where one person speaks and not everyone shares the language, or hears it well. Pick the one closest to yours. How to connect the sound in each

Seminar or workshop

Twenty to forty people around tables, one trainer, a laptop already plugged into the projector.

Clip-on mic→Laptop→Phones

How it is set up: Open the invite link on the trainer's laptop and put the QR code on the first slide. A clip-on or USB microphone works better than the laptop's own.
What people get: The trainer's words in their own language, on their own phone, as they are spoken. Project names and partner names stay exactly as said.
Session length: One invite of 45 or 90 minutes per module.

Live course or lecture

A lecture hall, the same teacher every week, and some students who are still learning the language or who hear poorly.

Lectern mic→Laptop→Screen view+Phones

How it is set up: One invite per lecture. Put the screen view beside the slides, in the language of the lecture, so the whole hall can read what was just said.
What people get: Students follow in their own language and save the text for their notes before it is deleted from the server.
Session length: 45 to 90 minutes, to match the timetable.

Conference talk

A stage, a sound system and a technician. Guests from several countries in the hall.

Stage mics→Mixing desk→Laptop→Phones+Side screen

How it is set up: Ask the technician for a feed from the mixing desk into the laptop, so the app hears exactly what the hall hears. Show the QR code on the holding slide and on the screen view.
What people get: Each guest picks one of 15 languages. Nobody queues for a headset or hands one back at the door.
Session length: One invite per talk or per block of the programme.

Panel or round table

Four or five people, a moderator, and questions from the floor.

Table mics→Laptop→Phones

How it is set up: One sound feed for the whole table. The app tells the voices apart and labels them. When a guest answers in another language, switch the spoken language with one click.
What people get: The discussion with who said what, in their own language.
Session length: 60 or 90 minutes.

Public meeting

A council session, a parents' evening, the first meeting of a project with partners from abroad.

Room mic→Laptop→Phones

How it is set up: Nothing to install and nothing to hand out. The code goes on the wall or on the agenda.
What people get: Residents and partners follow in their own language without giving a name or an email.
Afterwards: The transcript is deleted from the server 15 minutes after the meeting ends.

How it works

Three steps, all in a browser. The speaker needs a laptop and a microphone; listeners need a phone.

Open your invite link
Choose the language you will speak, add the names and terms that must not be mangled, and start. The link is yours for the length of the session.
Show the code
A QR code on a slide, or the screen view on the projector.
Speak
Each sentence appears while you say it, in every listener's language.

What you get

The same in every session, whatever the room.

Live textListeners watch the sentence grow, then settle into its final form.
A second readSeconds later an AI reviser mends punctuation and misheard names. A word it adds is marked, so nobody takes it for heard.
15 languagesEach listener picks their own and can change it at any moment.
Names keptPeople, places and acronyms you list are heard correctly and never translated.
Screen viewLarge captions for a projector or a side screen, with the code to join.
Who said whatVoices are told apart and labelled, with the speaker's name when it is given.
Nothing to installIt runs in the browser of any phone, light or dark.
Then it is goneDownload the transcript, as heard or in clean paragraphs; 15 minutes after the end it is deleted from the server. What happens to the words

The room answers back

Reading is half of it. On the same code, people in the room can ask, speak, vote and say how it went. The moderator decides when.

Ask in writing

A listener writes in the language they read in. The moderator sees the question translated and reads it out, so the answer is for everyone.

On the phone: A button appears when the moderator opens questions. Up to 280 characters, with a name only if they want to give one.
For the moderator: The question in the language being spoken, the original underneath, and a tick for when it is answered.
Afterwards: Questions are added to the transcript the organiser downloads.

Speak from the phone

A listener asks to speak. When the moderator gives them the floor, their phone is the microphone and the room's loudspeakers carry their voice.

On the phone: “Ask to speak”, then a red bar while they are live. They stop when they are done, or the moderator takes the floor back.
For the moderator: Who is waiting, one button to give the floor, and a volume control.
For everyone else: What is said is written down and translated like the rest of the talk, under the listener's name.

Answer a poll

The moderator writes a question with two to five answers and launches it. Each listener sees it in their own language and picks one.

On the phone: The poll comes up by itself. After answering, the listener watches the room's answer build up.
In the room: The same bars on the projector, live, next to the code to join.
Afterwards: The results are added to the transcript the organiser downloads.

Leave feedback

When the session ends, each phone asks for a rating from 1 to 5 and a few words. No name.

On the phone: It comes up by itself at the end. The moderator can open it a few minutes earlier, while people are still in the room.
For the organiser: The average, how the ratings spread, the comments translated, and a CSV file for the project report.
Kept: Fifteen minutes on our server, like everything else. Download it before then.

What it does not do

Better to read this here than to find out in front of a full room.

It does not speak. Listeners read. There is no translated audio.
It is machine translation. It gets things wrong, most often the names it was not told about.
It needs the internet, on the speaker's laptop and on every phone.
Speaking from a phone depends on the network. Some mobile networks do not let a phone reach the speaker's laptop. Writing the question always works.
It hears what the microphone hears. A laptop at the back of a hall is not enough. Sound for your room

Try it in your room

A real event starts from an invite link. Tell us what the event is, how long it runs and which language is spoken, and we will send you one.

contact@socialit.ro

Ask for an invite