Treble Secures $18 Million to Enhance Icelandic Voice Simulation Platform

Voice AI is rapidly becoming a leading sector in the field of artificial intelligence, attracting significant investments aimed at diverse applications. These range from automating customer service and sales interactions to pioneering technologies such as AI-driven meeting notetakers and smart glasses that leverage voice as a primary interface.

In this dynamic landscape, AI laboratories are consistently introducing new models, while hardware manufacturers strive to enhance user experiences with their devices. However, the burgeoning technological advancements require robust testing and feedback mechanisms for continual improvement.

Treble, a startup based in Iceland, is positioning itself at the forefront of the voice AI industry by offering a unique simulation platform tailored for model developers, robotics firms, and consumer electronics manufacturers. Founded in 2020 by acoustic engineers Finnur Pind and Jesper Pedersen, Treble has recently raised $18 million in an extension of its Series A funding, led by Paladin Capital Group, with contributions from existing investors such as KOMPAS VC, Frumtak Ventures, EIC, and Omega ehf. This fresh influx of capital brings Treble’s total funding to over $40 million, having previously secured $12 million in 2024. Notably, the startup counts industry giants like Amazon and Logitech among its clientele.

Treble specializes in several verticals related to simulation and data analytics. For companies focusing on voice AI, it offers a synthetic data generation platform that supports speech enhancement, noise suppression, and model training. Additionally, Treble evaluates voice AI models under varying conditions to furnish laboratories with essential feedback. Earlier this year, the company teamed up with Hugging Face to establish a benchmark for speech recognition models across a variety of realistic conditions.

According to Pind, “Audio AI is fundamentally a data challenge, presenting significant opportunities for the next generation of models and hardware. Traditionally, sound-related AI has relied on recordings and datasets sourced from the internet. We believe that accurate physics simulation can serve as a viable alternative in generating sound data.”

Moreover, Treble is keenly focused on hardware design and testing from the perspective of voice functionality. It collaborates with headphone and speaker manufacturers for virtual prototypes to help them analyze sound quality. The startup is also advancing its efforts in simulating testing for smart glasses and AI devices, including smart speakers that can interpret voice commands based on their spatial positioning.

Pind expressed enthusiasm for Treble’s work on wearable devices aimed at enhancing hearing capabilities. “I’m particularly excited about the next generation of products like headphones and smart glasses featuring superhuman hearing. Imagine being in a noisy restaurant and wishing to concentrate solely on nearby conversations or muting background noise at a seminar,” he remarked.

Looking ahead, Treble intends to deepen its focus on the physical AI domain, targeting sectors including robotics, automotive, and drone technology to facilitate sound-based functionalities through comprehensive testing and simulation.

Francois Ruether, VP of Paladin Capital Group, emphasized Treble’s distinctive platform that simulates various models and devices, asserting its growing importance as the technology expands into new areas. “As more products rely on sound interpretation, this underlying infrastructure will become increasingly invaluable across voice AI, wearables, robotics, and physical AI,” Ruether added. He highlighted that customers retain ownership of their models and development processes while leveraging a shared foundational infrastructure.