Scribe

GreenKey Scribe

Artifical Intelligence that understands financial language and enables powerful voice workflows




DOCUMENTATION

Start out benchmarking Scribe's accuracy, then customize it with your own data and deploy it in a real-time workflow. All of our services can be deployed on-premise on Docker compatible machines.


VOICE TRANSCRIPTION SERVER (SVTSERVER)

Transcribe audio files and obtain rich metadata, then teach the engine new words and phrases.

Read the docs >>
REAL-TIME DICTATION SERVER (SQCSERVER)

Dictate in real-time to populate forms, create alexa-like skills, or obtain quick general transcription from a voice stream.

Read the docs >>
ACOUSTIC AND LANGUAGE MODEL TRAINING (SCRIBETRAIN)

Extend Scribe's speech recognition model with new audio or text data to teach it new accents, voices, or tribal lexicons.

Read the docs >>
CALL BOT (CALLTRANSCRIBER)

Scribe's Call Bot uses a combination of microservices to help you transcribe SIP streams in real-time or at a delay.

Read the docs >>
DISCOVERY ENGINE (DISCOVERY)

Scribe's Discovery engine has powerful, customizable NLP to search for key phrases or create voice skills.

Read the docs >>



EXAMPLE WORKFLOWS

Transcribe Voice Streams for Compliance

Scribe's Voice Transcription engine creates time-stamped, searachble, speaker separated transcripts from audio files.



DIAL SCRIBE

Call Scribe's SIP endpoint directly or merge it into your calls.

SCRIBE CALL BOT

Scribe's Call Bot does the heavy lifting on handling SIP connections and broadcasting transcription over a websocket.

SCRIBE VOICE TRANSCRIPTION

Scribe's SVTServer will transcribe the stream near real-time.

TRANSCRIPT AND INSIGHTS

JSON formatted with additional metadata will be saved as well as broadcasted over a websocket relay.


Real-Time Workflows

Populate your own application trade ticket via voice, create Alexa-like skills, or dictate notes with our easy to setup Javascript API.



REAL-TIME AUDIO

Websocket API via computer microphone or SIP via the Call Bot

SCRIBE REAL-TIME DICTATION

Scribe's SQCServer will transcribe the stream in real-time and extract meta-data.

SCRIBE WEBSOCKET RELAY

The websocket relay server will forward the response to any backend capture system.

CAPTURED METADATA

JSON payload with metadata delivered to your application front-end.



Need access credentials?

Contact Us