Voice artificial intelligence has rapidly become one of the most competitive and heavily funded segments of the broader AI industry. Investors have collectively directed billions of dollars into companies building voice-driven applications, from automated customer support and outbound sales tools to AI-powered meeting assistants and smart glasses that rely on voice as the primary means of interaction.

At the same time, AI laboratories are accelerating the pace at which they release new models, while consumer hardware manufacturers race to deliver the most seamless voice interaction experiences on their devices. Underpinning all of this activity is a fundamental need that often goes unnoticed: rigorous testing and a reliable feedback mechanism to continuously improve how these systems perform in real-world acoustic environments.

A startup headquartered in Iceland is working to fill that gap. Treble has built a simulation platform designed to serve model developers, robotics companies, and consumer hardware manufacturers by giving them the infrastructure they need to test and refine voice-related technology before it reaches end users.

Funding Round and Company Background

Treble was founded in 2020 by acoustic engineers Finnur Pind and Jesper Pedersen. The company has now announced an $18 million extension to its Series A funding round, with the investment led by Paladin Capital Group. Existing backers KOMPAS VC, Frumtak Ventures, EIC, and Omega ehf also participated in the round. This latest tranche follows a $12 million investment the company received in 2024, bringing its total capital raised to more than $40 million. Among its current customer base, Treble counts major technology companies including Amazon and Logitech.

What the Platform Does

Treble's business is organized around several distinct areas tied to simulation and data. For companies building voice AI products, the startup offers a synthetic data generation platform that can be used for tasks such as speech enhancement, noise suppression, and the training of machine learning models. The platform also evaluates how voice AI models perform under a range of different acoustic conditions, feeding that performance data back to the developers for refinement.

Earlier this year, Treble entered into a partnership with Hugging Face to release a benchmark tool designed to measure the performance of speech recognition models across varied and realistic listening environments. The collaboration reflects growing recognition within the AI community that model evaluation needs to account for the full complexity of real-world sound conditions, not just controlled laboratory settings.

Co-founder Finnur Pind described the core problem Treble is trying to solve as fundamentally a data challenge. He noted that virtually all sound-related AI developed to date has been built using recordings and data gathered from the internet, and argued that physics-based simulation offers a more accurate and scalable alternative for generating training data.

"Audio AI is really a data challenge, and this is where the most opportunities to enable next-generation models and hardware lie. To date, pretty much all sound-related AI has been made from recordings and data scraped from the internet. We believe that accurate physics simulation can be an alternative way to create data for sound," Pind said.

Hardware Design and Consumer Device Testing

Beyond software and model evaluation, Treble also works directly with hardware manufacturers on voice-related design and testing. The company collaborates with headphone and speaker makers to support virtual prototyping, helping them understand how a product will sound before a physical prototype is produced. In practical terms, this might involve simulating how a smart speaker interprets voice commands depending on where the speaker is physically positioned in a room.

More recently, Treble has extended its simulation capabilities into the testing of smart glasses and other AI-enabled devices, a segment that has attracted significant interest from both established technology companies and newer entrants. Pind expressed particular enthusiasm for the potential of next-generation wearables to enhance how people hear in challenging acoustic environments.

"I'm really excited about the next generation of these devices like headphones and smart glasses that can enable 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.

Expanding into Physical AI

Looking ahead, Treble is directing more of its attention toward what is broadly referred to as physical AI - a category that includes robotics, automotive systems, and drone technology. The company intends to support these industries by providing simulation and testing services that enable sound-based functions, such as helping a robot interpret audio cues from its surrounding environment or allowing a vehicle's systems to respond accurately to voice commands in noisy conditions.

Francois Ruether, Vice President at Paladin Capital Group, explained the investment thesis behind backing Treble. He pointed to the platform's ability to simulate the behavior of different models and devices as a distinguishing feature, and argued that its strategic value will grow as the platform becomes embedded in more product categories and industries.

"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.

Treble's positioning reflects a broader shift in how the technology industry is thinking about the infrastructure required to support AI development. Rather than competing directly in the crowded space of AI model creation, the company is targeting the layer beneath - the testing, simulation, and data generation tools that model makers and hardware developers need to build reliable, high-performing products. As voice and audio AI continue to expand into new form factors and environments, that infrastructure layer is likely to attract increasing attention from both investors and industry partners.