About ML Structure

1. Requirment
   : All functions are configured and executed on aws server in aws lambda, so nothing is needed on the ML side.
   : Certifications for each lambda function.
   Go to
   ./api_server/aws_lambda_file/kpmg_transcribe/setting.json, ./api_server/aws_lambda_file/kpmg_comprehend/setting.json

2. AWS Key File
   Go to
   ./api_server/new_user_credentials.xlsx.
   But this key will not required because we set up access range to public

3. ML flow
   Take mp4 as input from the web and upload it to the kpmg-input-video s3 bucket. When the upload finishes, Amazon lambda is triggered and Amazon transcribe is called. The called Amazon transcribe uploads the data containing transcrip to the kpmg-output-text s3 bucket in json format when the execution is complete. Similarly, when the upload finishes, Amazon lambda is triggered and Amazon comprehend is called. Called Amazon comprehend upload key-pars and their time-lines to the kpmg-output-text s3 bucket in json format when the execution is complete. Call the uploaded json file on the web.

4. ML Structure Image