Imagine you are a Formula One driver hurtling down a race track at 200 miles per hour. Your engineer comes on the radio with a critical message, but you cannot make it out. There is no time to ask for a repeat with the race, and your life, on the line. This is one problem the Norwegian startup Hance is solving with an impressively small and fast piece of audio-processing software. The technology has already attracted customers like Intel and Riedel Communications, the official radio supplier to Formula One.
Hance is one of 200 startups selected to showcase its technology at TechCrunch Disrupt 2025. The event runs from October 27 through 29 at the Moscone Center in San Francisco. The company has around ten employees and brings a wealth of audio industry experience. This includes co-founder Stian Aagedal, who is also the CEO of audio editing software company Acon Digital, and Peder Jørgensen, who runs the sound effects library Soundly.
With the artificial intelligence boom, the Hance team saw an opportunity to leverage new technologies throughout the audio-processing pipeline, especially in noise reduction and isolation. A few years ago they began training their own models using Soundly’s high-quality recordings. These recordings included everything from the roar of Formula One cars to the crack and rumble of Icelandic volcanoes.
Since then, they have shrunk the Hance processing models to just 242 kB. This small size allows the models to run on a device instead of in the cloud, saving both time and energy. Hance states these models can separate sounds, remove noise, echo, and reverb, and enhance speech clarity with a latency of just 10 milliseconds.
While other companies offer similar audio processing software, Hance’s tiny and energy-efficient models can process audio on devices of all sizes in real time. This makes the technology ideal for the radios Riedel sells to Formula One and FIFA, and also attractive for law enforcement and defense applications, according to CEO Joote Hika.
Hika sees opportunity for Hance’s audio processing to expand in many new directions, especially now that Intel is a partner. Hance has been working with the technology giant to adapt its models for different versions of Intel chips, including the latest neural processing units. The startup is also in talks with other chipmakers and an undisclosed smartphone manufacturer.
Hika noted that these professional partnerships are non-exclusive and will likely last for several years. This is good for the startup’s ability to scale, but he said Hance must continue developing rapidly to stay ahead of the competition. The company recently hired its first chief commercial officer, but Hika expects Hance will remain heavily focused on research and development. The company will prioritize hiring AI-capable workers to stay lean.
Hika stated that while they currently have an advantage over competitors, they must maintain that lead through fast-paced development.