One of the best features of Slack or any other communication tool for teams is history, and being able to search all messages to find the right information. We use Slack daily in our company, and with the latest Add-On on voice notes, we though about adding some way to use voice notes with the possibility to search in all voice notes messages too.
What it does
The application allows the user to upload or record an audio message, then, using Watson API for Speech-To-Text we get the audio transcript + information about the time of each word. We store this information in ElasticSearch to have a fast, scalable and robust way to search through all this messages. Voice notes are stored in Amazon S3 and the application allow the user to search, find the second of audio they need and listen to it.
How we built it
The application is build using node.js for the recording or audio uploading screen, this application then submit the audio to Amazon S3 and index the transcript information into Elastic Search through a Ruby on Rails API.
What's next for Voice Tracker
Develop plugins or integrations with IM applications that can use this technology