Vatis Tech builds a speech-to-text transcription platform that emphasizes high accuracy for Romanian and English by training models on multiple domain-specific datasets (media, medical, phone calls, legal, ads, noisy audio). The company reports Romanian-language accuracy above 95% across several domains and roughly 90% average accuracy for general English speech. Its platform has transcribed over 3.1 million notes and serves clients in media, government and banking, including Unicredit and Klarmedia. The team trains models on both public datasets and client-specific data to improve robustness for particular use cases. Planned product work focuses on developing domain-specific transcription models (priority: economic field, also law and medicine) and area-specific solutions such as call-center models. Investors cite the team’s rapid product development and client traction as the rationale for continued support. Vatis Tech builds an AI-powered online speech recognition platform and API that supports multiple languages and industry verticals. The company aims to enable users to create custom speech recognition models with minimal resources using its infrastructure. It focuses on European languages—including French, Italian, German, Dutch and Polish—and started with Romanian. Vatis Tech’s algorithms are specialized for sectors such as media, finance, technology and medical, and can be deployed in the cloud or on premise. The service is available online and developers can integrate the API with dedicated support and pricing. In tests reported by the company, Vatis Tech’s transcription accuracy exceeded 94%, compared with competitors’ 84%, and it has recently closed a funding round of over €200K.

Total raised

$1M

Funding rounds

2

Latest round

Seed

Latest activity

Feb 2023