GTranscription is a commercial solution that converts FreePBX (Asterisk) Call Recordings into transcriptions, usually overnight, and optionally provides a customisable summarisation, analysis and further processing. An example of further processing, is to analyse the given transcripts (for example a list of sales calls) and score the interaction, rate the agent, and ensure the required compliance statements were included. The results of this feed into further systems for Q&A etc.
The GTranscription Module for FreePBX provides semantic search of transcriptions, with an option to download the call recording, transcription or JSON (including speaker identification & isolation etc) and is designed as a ‘search’ facility to quickly find call recordings with specific words, phrases (or not).
Nicely done. What is the pricing on this or is it open source? I notice on your website you state
“ When it comes to telephony, there really is only one platform that dominates, FreePBX.
GEN Actively contributes to the project, and releases code and documentation to the community, because we believe in the future of FreePBX and PBXact.”
Where can we see all this code, modules and improvements you have contributed to the project?
From the questions that get asked on the Asterisk forum, the market demand seems to be for “real time”, separated speaker transcription, so I wonder if this product is a couple of years too late.
(Personally it seems to me that doing real time, separated, speaker, transcriptions wastes the power of current AI, because the AI doesn’t have access to the full context of the segment it is transcribing. However, it seems to be what everyone is asking for.)
By separated speaker I mean that they want to process inbound and outbound speech directions independently of each other.
