Treble’s voice AI simulation startup raises $18M, pushing total funding past $40M

2 days ago 29

Voice AI has become one of the busiest corners of the tech industry, and the money keeps flowing toward it. Amid that rush, a small Icelandic company is betting that the real bottleneck isn’t the models themselves but the data used to train and test them. Treble, a startup building a voice AI simulation platform, has just raised $18 million to push that idea further.

Key takeaways

Led by Paladin Capital Group, Treble secured an additional $18 million as part of an extension to its Series A round.
The company’s total funding now exceeds $40 million, following a $12 million investment in 2024.
Founded in 2020 by Finnur Pind and Jesper Pedersen, Treble counts Amazon and Logitech among its customers.
The platform generates synthetic data for speech enhancement, noise suppression, and model training, and partnered with Hugging Face on a speech recognition benchmark.
Treble plans to expand its acoustic testing work into robotics, automotive, and drones.

Treble secures $18 million extension in Series A funding

Treble’s new capital comes as an extension of its Series A round, led by Paladin Capital Group. Existing backers KOMPAS VC, Frumtak Ventures, EIC, and Omega ehf also took part in the round, according to TechCrunch. That continued support from prior investors suggests confidence in the company’s direction rather than a one-off bet from a new lead.

Funding led by Paladin Capital Group

Paladin’s involvement carries some weight beyond the check size. Francois Ruether, VP of Paladin Capital Group, told TechCrunch that Treble’s simulation platform stands apart because it can model both AI systems and physical devices at once, and that its relevance should grow the more industries adopt it. “Our thesis is that, as more products depend on understanding sound, this infrastructure becomes increasingly valuable across voice AI, wearables, robotics, and physical AI,” Ruether said. He added that “customers retain ownership of their models, products, and development workflows, while benefiting from a shared foundation of a simulation-native acoustic infrastructure layer.”

Total funding exceeds $40 million

The fresh $18 million follows a $12 million investment the company received in 2024, pushing Treble’s total raised to date past $40 million. That trajectory places the startup among a small group of specialized voice-infrastructure companies attracting sustained institutional backing rather than a single splash of venture cash.

Company background and founding

Treble was founded in 2020 by acoustic engineers Finnur Pind and Jesper Pedersen, and it’s based in Iceland — a detail that sets it apart from the Silicon Valley-heavy voice AI crowd. The company’s pitch is built on physics rather than pure machine learning: instead of relying only on recordings scraped from the internet, Treble argues that accurate physics-based simulation can generate the sound data that next-generation AI models and hardware need.

“Audio AI is really a data challenge, and this is where the most opportunities to enable next-generation models and hardware lie,” Pind told TechCrunch. “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.”

Treble’s voice AI simulation platform and applications

At its core, Treble sells a synthetic data generation platform aimed at voice AI companies, and the value proposition is straightforward: better training data means better-performing models. This matters because most voice systems today still lean on recorded audio scraped from wherever it can be found, which limits how well models generalize to messy, real-world acoustic environments.

Synthetic data generation for voice AI

Treble’s platform produces synthetic voice data that companies can apply to speech enhancement, noise suppression, and model training. Rather than depending solely on existing audio recordings, the system generates simulated acoustic scenarios that can be tailored to specific testing needs.

Model evaluation and feedback

Beyond generating data, Treble also evaluates voice AI models under a range of conditions and feeds that analysis back to the labs building them. Amazon and Logitech are already customers, giving the startup real-world validation from companies that build consumer-facing devices at scale.

Partnership with Hugging Face

Earlier this year, Treble partnered with Hugging Face to launch a benchmark for speech recognition models tested across different realistic conditions. That collaboration puts Treble’s testing methodology in front of a much broader developer community than it could reach on its own, and it signals an attempt to become a recognized standard for evaluating how speech models perform outside controlled lab settings.

Hardware design and physical AI expansion

Treble’s ambitions extend well past software. The company also works directly with hardware makers, using acoustic simulation to shape how physical products sound and perform before they’re built.

Virtual prototyping for headphones, speakers, and smart glasses

Treble partners with headphone and speaker companies on virtual prototyping, letting them understand how a product will sound before a physical unit exists. Its testing also extends to how well a smart speaker can pick up commands depending on where it’s placed in a room. More recently, the company has moved into simulation testing for smart glasses and other AI-enabled wearables.

Pind pointed to hearing enhancement as a particularly promising direction for that hardware work. “I’m really excited about the next generation of these devices like headphones and smart glasses that can enable a feature like superhuman hearing,” he said. “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.”

Plans to expand into robotics, automotive, and drones

Looking ahead, Treble intends to deepen its presence in what’s often called physical AI — robotics, automotive, and drone companies — helping those industries build sound-based functions through the same kind of testing and simulation it already offers voice AI labs and hardware makers.

This is where Treble’s strategy starts to look less like a niche voice-tech play and more like an infrastructure bet. If sound understanding becomes a standard requirement for robots navigating a room, cars detecting hazards, or drones responding to audio cues, a company that already builds physics-based acoustic simulation tools has a head start most competitors would need years to build from scratch.