Deepgram launches improved AI-based voice transcription for enterprises
Accurate voice transcription is important for enterprises, whether it's to ensure appropriate responses or create accurate records.
However some situations make this challenging to achieve -- where there are multiple speakers or noisy backgrounds, for example. With the launch today of Nova-3, its most advanced speech-to-text (STT) model to date, Deepgram is looking to offer greater accuracy along with self-service customization to tailor results for industry-specific needs.
It makes use of ‘latent space architecture’ to encode complex speech patterns into a highly efficient representation. This enables superior transcription accuracy, even in noisy or specialized settings.
It can also recognize domain-specific terminology for use in specialized fields like medical and legal transcription. It offers accurate numeric recognition too for retail, banking and finance while supporting real-time redaction of sensitive information for compliance and data privacy.
"Nova-3 represents a significant leap forward, extending the frontier of real-time accuracy while once again bending the cost curve -- two critical components for enterprise speech-to-speech use cases," says Scott Stephenson, CEO of Deepgram. "By integrating advanced architectural enhancements and extensive training across diverse datasets, we've developed a model that not only meets but exceeds the evolving needs of our clients across various industries."
In addition Nova-3 enables real-time transcription across multiple languages -- the first model of its kind to do so -- making it suitable for emergency response, global customer service, and multilingual operations.
Self-service customization lets users fine-tune the model for specialized domains without requiring deep expertise in machine learning. Many conventional models require expensive and time-consuming expert-led customization, delaying deployment and increasing costs. Using Keyterm Prompting in Nova-3 developers can improve transcription accuracy by optimizing up to 100 key terms without waiting for extensive model retraining or customization cycles.
You can find out more on the Deepgram site.
Image Credit: SergeyNivens/depositphotos.com