Rocket.Chat users have an option of sending audio files or live audio messages to a room or a chat. Speech To Text app utilises the capabilities of Rocket.Chat.Apps-Engine and allows the user to be able to not only transcribe the audio files but also save the transcript as metadata. This whole functionality is packaged into a Rocket.Chat.App, which can be configured to use the transcription provider of user's choice.
-
Transcribes the audio files on server side & store the transcript metadata on demand.
-
Provides the User and option to choose from a variety of API providers for trascription. Currently supported providers are:
- Assembly AI.
- Mcirosoft Cognitive Services .
- & Rev AI.
All of the listed deliverables were completed within the GSoC period. YAY !! 🎉
-
Clone the repository and install the dependencies :
git clone https://github.com/RocketChat/Apps.SpeechToText.git
cd SpeechToText
npm install
-
Make Sure you have Rocket.Chat server running on your localhost & you have Rocket.Chat.Apps-cli installed, if not :
npm install -g @rocket.chat/apps-cli
-
You need to enable ==Apps development mode== in
Administration > General
-
Now, install the app:
rc-apps deploy --url http://localhost:3000 --username "your username" --password "your password"
-
Once successfully installed you'll be able to see the app in
Administration > Apps
. -
Enable ==Use JWT for file Upload== setting from
Administration > General
and add aJWT
secret. -
Provide the same
JWT
as in fileUpload and select your provider. -
Create a
Tunnel
to thelocalhost
and provide it's URL into the App settings (You need to provide this Tunnel URL for the app to work ). -
ENJOY 😉