Voice AI has rapidly become one of the most dynamic areas in artificial intelligence, attracting billions of dollars in investment across a wide range of applications—from automating customer support and sales calls to building meeting notetakers and AI-powered smart glasses where voice serves as the primary interface. At the same time, AI research labs are rolling out new models at a rapid pace, and hardware manufacturers are competing to deliver the best possible consumer experience with their devices. This surge of innovation demands rigorous testing and continuous feedback loops to drive improvement.

An Iceland-based startup called Treble is carving out a central role in the voice AI ecosystem by building a simulation platform designed to serve model developers, robotics companies, and consumer hardware makers alike. Founded in 2020 by acoustic engineers Finnur Pind and Jesper Pedersen, the company has secured $18 million in an extension of its Series A round, led by Paladin Capital Group, with backing from existing investors KOMPAS VC, Frumtak Ventures, EIC, and Omega ehf. This follows a $12 million investment in 2024, pushing Treble's total funding to over $40 million. Among its customers are Amazon and Logitech.

Treble operates across several verticals tied to simulation and data. For voice AI companies, it offers a synthetic data generation platform that supports speech enhancement, noise suppression, and model training. The company also assesses voice AI models under varying conditions to give labs actionable feedback. Earlier this year, Treble teamed up with Hugging Face to introduce a benchmark for speech recognition models tested across diverse, realistic scenarios.

"Audio AI is really a data challenge, and this is where the most opportunities to enable next-generation models and hardware lie," noted the company. "To date, pretty much all sound-related AI has been made from recordings and data scraped from the internet."

Beyond data and model evaluation, Treble concentrates on hardware design and testing from an audio perspective. It collaborates with headphone and speaker manufacturers on virtual prototyping, helping them predict how their products will sound. The platform can also evaluate how well a smart speaker interprets commands depending on where it's positioned. More recently, Treble has expanded into simulation testing for smart glasses and AI-driven devices.

Pind expressed enthusiasm about working on wearables capable of enhancing hearing for users. "I'm really excited about the next generation of these devices like headphones and smart glasses that can enable [a feature like] superhuman hearing. That's an area where you can really just hear better in challenging acoustic environments. Maybe you are in a restaurant, and you only want to hear people within two meters of range, or you are in a seminar, and want to mute people around you," he said.

Treble is looking to deepen its involvement in the physical AI space, including robotics, automotive, and drone companies, enabling sound-based capabilities through testing and simulation. "Our thesis is that, as more products depend on understanding sound, this infrastructure becomes increasingly valuable across voice AI, wearables, robotics, and physical AI. Customers retain ownership of their models, products, and development workflows, while benefiting from a shared foundation of a simulation-native acoustic infrastructure layer," Ruether said.