Iceland-Based Treble Raises $18 Million to Build Voice AI Simulation Platform

Iceland-based startup Treble has raised $18 million to expand its voice AI simulation platform, which helps companies develop and test audio-based artificial intelligence systems for applications including consumer devices, robotics, and smart hardware.

Paladin Capital Group led the funding extension, with participation from existing investors including KOMPAS VC, Frumtak Ventures, EIC, and Omega ehf. The latest round brings Treble’s total funding to more than $40 million.

Building simulation tools for voice AI

Founded in 2020 by acoustic engineers Finnur Pind and Jesper Pedersen, Treble is building a simulation platform designed to help AI companies create synthetic audio data, evaluate voice models, and improve speech performance in different environments.

The company provides tools for speech enhancement, noise suppression, and AI model training. Earlier this year, Treble partnered with Hugging Face to launch a benchmark for testing speech recognition models under realistic conditions.

Treble says traditional audio AI systems have largely depended on recorded data, while physics-based simulation can provide another way to generate sound data for future models.

Expanding into wearables and physical AI

The startup also works with hardware companies to test how products such as headphones and speakers perform before production. Its technology can simulate how devices capture sound, understand commands, and operate in different acoustic environments.

Treble has recently expanded into testing for smart glasses and AI-powered devices, while also exploring applications in robotics, automotive systems, and drones.

The company has also worked on speech technology evaluation, including a speech recognition benchmark project with Hugging Face.

With the new funding, Treble plans to expand its platform as demand grows for voice-based AI systems and devices that rely on understanding sound.