Introducing Vatis Tech V7 Transcription Model
Adrian Ispas

Adrian Ispas

May 7, 2024

Vatis Tech Embarks on a New Era: Introducing Our V7 Transcription Model


Experience the Future of Speech Recognition Today

Try Vatis now, no credit card required.

At Vatis Tech, we are always on the frontier of technological advancements, striving for excellence and innovation. It's with great excitement that we announce a significant milestone in our journey: the launch of our V7 transcription model. This new iteration marks a leap forward in our commitment to delivering exceptional transcription accuracy and quality, particularly for the Romanian language model.

V7's Groundbreaking Improvements

The V7 model represents the culmination of our relentless efforts in refining and enhancing transcription capabilities. With over 2000 hours of training data, the V7 model sets a new standard for accuracy and performance. 

Here's a closer look at the transformative improvements we've achieved with V7:

Overall Word Error Rate (WER): We've managed to reduce the overall WER to 0.0583 from V6's 0.0605, enhancing the model's precision across all types of audio content.

Phone Calls: Special attention was given to phone call transcriptions, where V7 shows a remarkable improvement, boasting a WER of 0.0559 compared to V6's 0.0597. This 7% enhancement underscores our commitment to supporting clear and accurate communications.

Medical Transcriptions: In a significant leap forward, the V7 model dramatically improves the transcription accuracy of medical audios. With a WER of 0.0492, it represents a 38% improvement over V6's 0.0740. This advancement is a testament to our dedication to supporting healthcare professionals with reliable transcription services.

Tailored to Excellence: The Romanian Model

Moving from V6 to V7, our ambition extends to exceeding 95% accuracy rate, with a keen focus on crucial sectors such as phone conversations and medical recordings, demanding utmost precision. By refining the WER to 0.0583 and harnessing over 2000 hours of specialized Romanian speech datasets, V7 pioneers new frontiers in transcription accuracy.

Forward Together

The journey from V6 to V7 has been one of meticulous research, development, and community engagement. We believe that these advancements will not only enhance the user experience but also empower professionals across various sectors with more reliable tools for documentation and analysis.

Our vision for Vatis Tech is not just about advancing technology; it's about creating solutions that make a difference in people's lives. The V7 transcription model is a significant step toward this vision, offering improvements that matter to you.

We invite you to experience the enhanced capabilities of our V7 model. Your feedback and insights are invaluable to us as we continue to refine and evolve our technology. Together, let's redefine the possibilities of speech-to-text technology.

Thank you for being part of our journey. Here's to many more milestones ahead.

Continue Reading

Experience the Future of Speech Recognition Today

Try Vatis now, no credit card required.

Waveform visual