OT-AudioTranscription (5)

[frlzjosh]

WHAT TO DO:

1. Create a microservice in flask that will consume from the 'outgoingAudioEvent' topic.
2. You can base your consumer from this location (https://github.com/Off-Top-App/off-top-python/blob/master/backend-services/Services/Kafka/Consumer.py)
3. From this object you are given a string URL that is a reference to an audio file we have uploaded on GCP storage bucket, you will need to retrieve this file in Python
4. Then when you have this file you will need to pass this file Mozilla Deep Speech API to transcribe it into text (Code is created but you will need to modify it so you can pass the file path:(https://github.com/Off-Top-App/off-top-python/blob/e58676244359c7c89bd8e4da3a9ab0746ba4214f/mozilla-deepspeech/mds.py)
5. When you're done you will need to send an object with the following attributes {user_id, transcribed_text, time_exported, topic, analyzed_at} into your Kafka Producer

RESOURCES:

1. How to set up Mozilla deep speech API on your machine: https://github.com/Off-Top-App/off-top-python

TASK IS DONE WHEN:

1. You can send a recording from flutter and see that this audio file gets transcribed in your microservice. As well as being able to send this data to your Producer without any errors.