ONDEWO Speech-to-Text (S2T) Platform

ONDEWO Speech-to-Text (S2T) Platform

Your spoken word gets transcribed into text with state-of-the-art accuracy.

  • Human voice gets automatically transcribed
  • Applications, devices and processes can be voice-controlled
  • Customer use case specific speech-to-text models for even the most complex vocabulary and phrases
Benefits

What you gain

ONDEWO Speech-2-Text (S2T) Platform is a software platform for enterprises to transcribe human voice in form of audio to text. Our unique approach to speech-to-text unlocks many benefits for companies with a more complex vocabulary set or phrases to recognize such as achieve a highly accurate transcription of human voice for a specific use case and optimize transcription for speed.

Our main USP here is that we are able to develop highly customer-specific speech-to-text models with incredible fast transcription speeds for multiple languages based on our pre-trained general and industry specific speech-to-text models.

Highest accuracy & speed

With our newest deep learning algorithms, the speech recognition is highly accurate (85-95%) with a 40-60x faster transcription speed than cloud providers.

Multiple languages & dialects

ONDEWO specializes in ensuring that the AI-Agent can understand various languages, and also dialects and non-native speakers. Need a new language or specific dialect, we don’t have yet – no problem, we teach our AI within 3-4 weeks.

Recognizes voices from various channels

It recognizes and transcribes voices coming through your different channels like the telephone (8kHz), mobile phone, laptop or tablet (16kHz).

Demo

Convert what you say into text

Choose whether you want to upload an audio file or say something live.

01

Select real-time audio stream transcription and click the microphone and say something.

02

The AI will do the rest and transcribe it for you.

03

You can then copy the text.

Important Features

What makes it special

Supports multiple languages including dialects

The Speech-to-Text algorithms and models come out of the box with supports 10 languages (German, English, Spanish, French, …) 

High speed speech-to-text transcriptions

The AI can support the hundreds of parallel transcriptions per installation with a 40-60x faster transcription speed than other cloud providers, e.g. enabling the system to have hundreds of phone call conversations in real time.

Use case-specific speech-to-text models

For complex topics with highly specialized vocabulary (e.g. product names, support ticket categories) we use cutting-edge transfer learning to train the speech-to-text algorithms and models for your specific use case resulting in significantly higher accuracy than general models from cloud providers. It allows the most different customers to use it, even those who otherwise are not understood by speech recognition.

Word error rates (WER) range from 10-15% on general models and less than 10% (human level) for customized specific speech-to-text models. 

Two recognition modes: “real time” and “batch file”

Real-time mode transcribes text while listening to audio streamed to the S2T platform (e.g. via web socket interface). This transcribed text is continuously updated and improved the longer the AI listens to the audio stream (“hypothesis of transcriptions updated”). Batch file mode transcribes audio to text based on a WAV audio file sent to the S2T Platform.

Various channels supported

The platform supports the transcription of human voice from various channels such as telephone (8 kHz) and mobile phone, laptop or tablet (16 kHz). We are currently working on building models for radio messages.

Quick and easy integration into products and services

All platform capabilities are easy to integrate into your products and services via ondewo-client libraries for various programming languages (e.g. Python, Nodejs, Angular, JavaScript etc.) and gRPC Remote Procedure Calls (“GRRPC”).

On-premise or cloud hosting

All features are available on both versions. The on-premise solution easily installs on all operating systems with standard Docker deployment environment on your own IT Infrastructure. Furthermore, it allows for a higher degree of data protection and control. With the cloud option, there is no maintenance needed as it gets provided by ONDEWO. The cloud option is billed on a usage basis. 

News

What else is new

Register here for the ONDEWO newsletter. You will be up to date, and you will always be the first to know about new products and solutions as well as other news.

Get Started

Get started right away!

What challenge is your company facing? An ONDEWO expert will be happy to support you in finding the best solution!

Become an ONDEWO partner

Get to know our Products

See all industry solutions