Your favorite tools. Sound for everyone.
From web to mobile, connect audio to your listener’s hearing profile. NABILA libraries bring integration to the languages and stacks you work with.
One idea. In your language.
Meet NABILA: a deep learning model we're building to support hearing-aid users inside the apps they already use. Because hearing is different for everyone.
The same starting point audiology has always had: a hearing test maps how you hear at every frequency. That chart is your audiogram.
Hearing-aid fitting has relied on trusted prescription formulas for decades. The best known is NAL-R: read each threshold on your audiogram, and prescribe back roughly half of the missing hearing, frequency by frequency. Everything NABILA does starts from this idea: your profile and the sound in front of you.
Every frequency matters. NABILA uses the hearing profile to adjust frequency levels, instead of turning everything up together.
Same audio · different hearing needs
Six thresholds guide the transformation
Your audiogram is the starting point
Audiogram in. Prescription levels out.
Levels informed by the audiogram
Different frequencies. Tailored levels.
The same audio goes in. The hearing profile guides the change. Switch profiles to explore the idea: personalization is more than louder sound.
A different emphasis on higher frequencies.
A visual exploration of personalizationFrom web to mobile, connect audio to your listener’s hearing profile. NABILA libraries bring integration to the languages and stacks you work with.
One idea. In your language.
Streaming sends small audio chunks for processing against the hearing profile and returns them in sequence. No waiting for the call to end or a complete file to upload.
Incoming chunks from your app
Chunks returned after processing
Scroll to follow each chunk through the journey.
Our service design follows one principle: your data serves your request only. Process audio to return a result, without retaining content or using it to train the model.
Temporary processing of audio and hearing profiles to produce the requested result.
No audio or hearing-profile storage after the request is a principle of the service design.
Request content stays outside model training and development data.
Leave your email and tell us how you would use NABILA — for yourself, or for an app that supports people with hearing aids.
Your invitation arrives by email before the public launch.