Iceland’s Treble Secures $18 Million Funding for Innovative Voice Simulation Technology
The voice AI sector is booming, drawing significant investments aimed at applications such as automating customer service, sales calls, and developing next-gen devices like AI smart glasses that utilize voice interaction.
As AI laboratories rapidly roll out new models, hardware manufacturers are focused on enhancing user interactions with various devices.
This influx of innovation requires rigorous testing and a mechanism for ongoing enhancement. An Icelandic startup, Treble, is striving to establish itself as a key player in the voice AI field by offering a simulation platform designed for model developers, robotics firms, and consumer electronics manufacturers.
Founded in 2020 by acoustic engineers Finnur Pind and Jesper Pedersen, Treble has successfully raised $18 million in an extension of its Series A funding, primarily led by Paladin Capital Group, and supported by existing investors like KOMPAS VC, Frumtak Ventures, EIC, and Omega ehf. In 2024, it secured an additional $12 million, bringing the total funds raised so far to over $40 million. The startup lists Amazon and Logitech among its clientele.
Treble operates across several sectors pertaining to simulation and data. For voice AI businesses, it offers a synthetic data generation platform suitable for applications like speech enhancement, noise suppression, and model training. Additionally, the company assesses voice AI models under various conditions to provide valuable feedback to development teams. Earlier this year, Treble collaborated with Hugging Face to establish benchmarks for speech recognition models tested in realistic scenarios.
“Audio AI primarily revolves around data challenges, presenting numerous opportunities for developing next-gen models and hardware. Traditionally, sound-related AI has relied on recordings and data sourced from the internet. We propose that precise physics simulation can create alternative data for audio applications,” Pind explained.
The startup also concentrates on hardware design and evaluation from a voice-centric viewpoint. For example, it collaborates with headphone and speaker manufacturers for virtual prototyping to determine how their products might perform acoustically. It can also analyze how well a smart speaker receives commands based on its positioning. Recently, Treble has begun offering simulation testing for smart glasses and AI devices.
Pind expressed enthusiasm for advancements in wearable technology that could significantly improve hearing capabilities for users.
“I’m genuinely excited about the next generation of devices like headphones and smart glasses that can provide features such as enhanced hearing. This technology could enable users to focus on conversations happening nearby, even in noisy environments like restaurants,” he noted.
Treble intends to further its endeavors in the physical AI landscape, including applications in robotics, automotive, and drone sectors, to foster sound-based functionalities through comprehensive testing and simulation.
Francois Ruether, VP at Paladin Capital Group, remarked that Treble’s platform, which simulates various models and devices, is exceptional; its significance is set to grow as its applications expand.
“Our belief is that as more products rely on their ability to perceive sound, the importance of this infrastructure will become increasingly critical in areas like voice AI, wearables, robotics, and physical AI. Clients maintain ownership of their models and development processes while benefiting from a shared simulation-based acoustic infrastructure,” Ruether concluded.
