Voice AI has emerged as one of the hottest areas in AI, with investors pouring billions of dollars into use cases ranging from automating customer support and sales calls to creating meeting note takers and developing AI smart glasses that use voice as the primary interaction surface.
Meanwhile, AI labs are rapidly releasing models and hardware manufacturers are trying to create the best experience for consumers to interact with their devices.
All of this new technology requires a feedback loop for testing and improvement. Iceland-based startup Treble aims to be at the heart of the voice AI industry by creating a simulation platform that can accommodate model makers, robotics companies, and consumer hardware manufacturers.
Founded in 2020 by acoustic engineers Finnur Pind and Jesper Pedersen, the company has raised $18 million in extended Series A funding led by Paladin Capital Group, with participation from existing investors KOMPAS VC, Frumtak Ventures, EIC, and Omega ehf. The company received an investment of $12 million in 2024, bringing its total funding to date to more than $40 million. The startup counts Amazon and Logitech as customers.
Treble has several areas related to simulation and data. For voice AI companies, it has a synthetic data generation platform that can be used for speech enhancement, noise suppression, and model training. It also evaluates voice AI models under different conditions and provides feedback to the lab. Earlier this year, the company partnered with Hugging Face to begin benchmarking its speech recognition model across a variety of realistic conditions.
“Audio AI is truly a data challenge, and this is where we have the best chance to enable the next generation of models and hardware. To date, nearly all sound-related AI has been created from recordings and data collected from the internet. We believe that accurate physical simulations can be an alternative way to create data for sound,” Pindo told TechCrunch.
The startup also focuses on hardware design and testing from an audio perspective. For example, we work with headphone and speaker companies to do virtual prototyping to help them understand how their products will sound. You can also test how your smart speaker can understand commands based on the speaker’s location. Recently, Treble has started offering simulation testing for smart glasses and AI devices.
Pind said the company is excited to work on developing wearables that can enhance users’ hearing.
“I’m really excited about the next generation of devices such as headphones and smart glasses that enable superhuman hearing (like capabilities). That’s the area where you can actually hear well in difficult acoustic environments. Maybe you’re in a restaurant and only want to hear people within two meters. Or maybe you’re in a seminar and want to mute the people around you,” he said.
Treble is increasing its focus on the physical AI space, including robotics, automotive, and drone companies, and aims to enable sound-based capabilities for these companies through testing and simulation.
Francois Ruether, a vice president at Paladin Capital Group, told TechCrunch that Treble’s platform, which simulates a variety of models and devices, stands out. As more areas become involved, the importance of the platform increases.
“Our argument is that as more products rely on understanding sound, this infrastructure will become increasingly valuable across voice AI, wearables, robotics, and physical AI. Customers can retain ownership of their models, products, and development workflows while benefiting from a shared foundation of simulation-native acoustic infrastructure layers,” said Ruether.
If you make a purchase through links in our articles, we may earn a small commission. This does not affect editorial independence.
