Berlin’s ai-coustics raises $1.6M with participation from Finnish FOV Ventures, to pioneer studio-quality speech in real-time

So far, Voice AI has mainly focused on automatic speech recognition, voice cloning and speech synthesis companies. Now, Berlin-based startup ai-coustics is developing a new category in the field by building the new standard of speech quality across all devices and applications.

The company has raised a €1.6 million seed funding round led by UK-based Connect Ventures, with participation from Inovia Capital and Finnish FOV Ventures. The round also includes notable angel investments from Michael Schmitt (venture partner at Earlybird-X), Jan Bohl (CFO of music production software Ableton), and Nigel Toon (co-founder and CEO of British unicorn Graphcore).

Even though we use our voices to communicate in numerous applications on a daily basis, audio and speech quality in media content is underwhelming. Phone calls, video conferencing, radio, podcasts, online lectures and TV series often lack clarity. While speech quality improvement in most real-time applications equals to removing background noise, this is only one way to increase intelligibility.

This is where ai-coustics comes in. Imagine every call, podcast or social media post sounding like it was produced in the acoustically perfect studio of a leading broadcaster. Ai-coustics is taking on low speech quality in digital communication by developing Generative Audio AI technology that goes far beyond noise suppression.

Ai-coustics was founded in 2021 at Technical University Berlin, in the capital of audio technology, by Corvin Jaedicke and Fabian Seipel. The founders were both Audio and Machine Learning engineers frustrated by the audio quality of online courses and tutorials.

Ai-coustics CEO Fabian Seipel comments, “AI technologies in smartphones and consumer devices have already revolutionised image and video processing. There is a huge opportunity for Voice AI to follow this trend, eliminating the need for expensive microphones, recording studios or post-production; and giving consumers access to high quality speech content across any medium. At ai-coustics, our ultimate goal is to set a new HD standard for audio quality.”

Sitar Teli, managing partner at Connect Ventures says, “Poor quality voice is something we experience on a daily basis, whether it’s while watching TV, speaking on the phone or on a conference call. Ai-coustics’ solution is a classic case of software eating the world – they’re replacing expensive audio components and equipment with software that can produce arguably higher quality voice output. The applications and implications of their technology are compelling, not just for existing use cases, but for new use cases they enable such as always-on augmented hearing. We’re excited to back Fabian, Corvin and Tim to build the best team in the world working on Voice AI.”

The startup has launched a suite of products including:
● A web application for content creators which enables perfect speech quality, offering a one-click solution for clear audio
● An API with manual control and fine-tuning options for broadcasters and media platforms to enable large-scale media enhancement
● An SDK to integrate high-quality speech into any kind of audio software or hardware application
● Underpinning these products is a sophisticated generative machine learning model that transforms any audio input to studio-level speech quality in real-time
● A self-built audio inference engine enables the tech to run on everyday consumer devices with limited computing and memory resources, making high-quality speech accessible for a variety of use cases: streaming, TVs or soundbars, transparency modes on headphones, significantly improved communication for hearing aid users, and a new level of voice clarity in online communication.

Unlike its voice cloning or text-to-speech cousins, ai-coustics’ approach preserves the characteristics and timbre of the speaker’s original voice. This ethical stance is important in an industry where concerns about digital impersonation and misuse are growing. Ai-coustics’ commitment to authenticity aligns with a growing demand for transparency and integrity in Voice AI.

The robust voice technology has global application – working across all languages, accents, gender groups and ages, effectively avoiding common pitfalls like hallucinatory audio effects and word changes in less common languages.

The seed funding will enable ai-coustics to expand their team and build partnerships across various sectors, enabling the business to deliver on its commitment to setting new standards in speech quality and improving user experiences across a range of devices and applications.

About ai-coustics
Berlin-based startup ai-coustics combats poor speech quality and intelligibility in audio communication and content creation with Generative Audio AI algorithms for everyone, across every device. Our technology empowers media creators and audio companies to incorporate a new standard of speech quality into their products and productions, improving the way we communicate and consume content. Read more.