Threll.ai
Voice Lab

Threll.ai launches Voice Lab

Where the voice becomes intelligence

BySveinung Totland4 min read
Four people in an audio lab, where a man in headphones sits at a large microphone and a colleague follows a sound wave on screen

A national volunteer effort to give AI a real, human voice

There are few things that say more about who we are than the way we speak. Our dialect reveals where we come from, the words we choose show who we are, and the rhythm of the language tells how we feel.

In encounters with artificial intelligence, all of this disappears. The voice becomes flat. The words lose their tone. AI may understand what you say – but not how you say it.

Therefore, we are launching Threll Voice Lab – an initiative to build the world's best voice AI for the Norwegian language.

A Norwegian language project – with a focus on Scandinavia

The Norwegian language is beautiful – but a nightmare for artificial intelligence. We have two official forms, hundreds of dialects, and significant variations in intonation, speed, and vocabulary. For an AI trained on English, Norwegian sounds like ten different languages at once.

Therefore, we start here – in Norway. We will train artificial intelligence to understand us as we actually speak. Once we master the Norwegian language, the technology will form the basis for expanding to the entire Scandinavian language area.

Our vision is to give artificial intelligence the ability to understand – and speak – Norwegian. Our beginning is to master all the dialects that make Norway unique.

From keyboard to voice

Artificial intelligence has learned to read and write. Now it must learn to speak and listen.

The transition from text-based AI to voice AI is enormous. A voice must interpret meaning through breathing, pauses, background noise, and emotions – not just letters. At Threll Voice Lab, we research and develop the technology that makes this possible:

  • Speech recognition (ASR) – that understands what is being said, even when the line is noisy.
  • Natural language understanding (NLU) – that interprets what you actually mean.
  • Dialog management and model optimization – that makes the conversation natural, human, and intelligent.
  • Text-to-speech (TTS) – that gives AI a voice that feels authentic, warm, and close.

Together, this makes up the technological core of Threll.ai's voice agents – Thrells.

Want to talk it through with someone who has done this before?

Fifteen minutes on a call. We will tell you honestly whether a voice agent fits your setup – and what has to be settled before it goes live.

Book Meeting (opens in a new window)+47 24 12 66 30

15 minutes · no obligation · pick your own time

The Threllers – the autonomous voices

Threll Voice Lab is not just a research project. It is the foundation for everything we build in Threll.ai.

Through Voice Lab we are developing voice agents – autonomous AIs that can hold meaningful conversations with humans, on the human's terms. They are called Voice Thrells, and the first ones are already being used in pilot projects:

  • Incoming Voice Thrells – that receive calls, assist customers, and conduct natural conversations.
  • Outgoing Voice Thrells – that make contact, build relationships, and create new business opportunities.

Both types are based on the same core: genuine language understanding. And it begins in Voice Lab – with your voice.

A technological leap for Norway

English dominates the world. Most AI models are trained on American English – and see Norwegian as a secondary language. The result? Norwegian businesses get voice technology that doesn't understand their customers. People get machines that speak unnaturally, and our language loses its place in digital development.

This is what we want to change at Threll.ai.

With Threll Voice Lab, we build language technology on a Norwegian foundation. We train models that understand Bokmål, Nynorsk, and all the dialects in between. We create datasets that reflect our culture – not translations from English. And we do it together, as a national volunteer effort.

For AI to understand Norwegians, AI must listen to Norwegians. And it starts here – in Norway.

The voices of the future

This is not just about technology. It's about preserving the way we speak. About making sure our voices don't disappear into the silence between the lines of an algorithm. About building artificial intelligence that understands us because it comes from us.

Therefore, we say:

  • Norwegian voices will be heard – in future technology.
  • Dialects are not noise – they are culture.
  • Language is not just communication – it is identity.
  • And artificial intelligence should not replace humans – it should enhance them.

This is our promise. This is Threll Voice Lab.

Join the volunteer effort

Give your voice to research. Let your dialect live on in future AI. Sign up for Threll Voice Lab

Related Articles

Continue reading more articles

See all articles