Whispp gives voice to throat cancer patients

The Dutch start-up Whispp took advantage of CES 2025 to present the new features of its eponymous application, designed to help people whose voices have been damaged (throat cancer, vocal cord damage, dysphonia, laryngeal paralysis… ) to communicate better orally.

It converts a hoarse, broken voice or even a whisper into a more natural voice using artificial intelligence techniques. This attachment to naturalness is its advantage over competing solutions. “It preserves the way I speak, it’s not about not just do speech-to-text then text-to-speech, because you lose a lot of information that way,” comments Akash Raj Komarlu, CTO and co-founder of Whispp.

Ultra-low latency and no need for the cloud

The big news this year (Whispp had already won an Innovation Award at CES last year) is that the models run with very low latency, between 60 and 170 milliseconds, and run locally on the smartphone. “Our ambition is to go down to 50 ms”says the CTO.

The app is available on iOS and Android, with a subscription model. Each new user gets 30 free call minutes, then pays 19.99 euros per month for 120 minutes. The technology works regardless of the language and is already available in France.

The ambition to be installed by default on smartphones

“Our current economic model is B2C, but the subscription is reimbursed by the government in the Netherlands. We are also in discussions with telephone operators in the United States and Japan for its integration into their services. Ultimately, we want license our technology to smartphone manufacturers so that it can be integrated directly into devices”explains Joris Castermans, co-founder and CEO of Whispp.

Note that Whispp can be useful even to people who have completely lost their voice. An audio message feature is available that converts text to voice and can even be trained on recordings of the user’s old voice to sound like them. “Our technology is also useful for those wishing to make calls discreetly, and is even compatible with bone conduction systems to make voice sound more natural”specifies Joris Castermans.