Amsterdam-based Reson8, a speech
technology company developing hyper-customised automatic speech recognition
(ASR), has raised €5 million in a pre-seed round led by Balderton Capital, with
participation from NP Hard.
The company aims to address
limitations in existing speech recognition systems, where performance often
declines across European languages, dialects, and domain-specific terminology.
Instead of relying on generic models, Reson8 is building acoustically precise
systems that can be adapted in real time using contextual data such as
documents, websites, and calendars. Its approach uses lightweight, modular
adapters to tailor recognition to specific conversations without increasing
latency.
Founded by Thomas Kluiters, Raoul
Ritter, and Jarno Verhagen, the team brings experience in high-performance
systems, AI, and speech technology. The company is focused on developing models
that adapt to how language is used in real-world settings, particularly in
professional environments where accuracy is critical.
Reson8’s technology is designed for
applications in sectors such as healthcare, legal, and customer support, where
specialised vocabulary and fast-paced communication can challenge traditional
systems. By prioritising precision and adaptability, the platform aims to
reduce errors in transcription and improve usability across different use
cases.
Generic speech models break the
moment you leave English, and customers increasingly want control over where
their data lives. We believe Europe can build its own speech infrastructure
that is customisable, accurate, and operated locally,
said Raoul Ritter,
co-founder of Reson8.
In contrast to providers reliant on
US-based infrastructure, Reson8 is building and operating its own European
infrastructure, offering greater control over data residency and compliance
with regional requirements. At launch, the platform will support more than 20
European languages.
The funding will be used to expand the
company’s infrastructure footprint in Europe, continue development of its
speech models and inference stack, and grow the team.