Hello,
My goal is to perform a recording with contact lens to redact voice/chat files and **never ** save the original files. For compliance reasons, I do not want this information in our systems if i can help it. It's not clear in the documentation, at least to me, that the original files in the S3 bucket from a recording will need to exist.
I was able to complete the post-analysis redaction and limit what an agent can see when viewing a contact. They are unable to listen/see the unredacted conversation and only have access to redacted information. Great! From the Connect web page, this all seems to work fantastically.
Now, to what is not so great. The S3 bucket still has the original file in the S3 bucket with the unredacted information. Yuk! I don't want to store any sensitive data, if possible, in my S3 bucket.
The documentation seems to be a bit confusing on this matter: (https://github.com/awsdocs/amazon-connect-admin-guide/blob/master/doc_source/enable-analytics.md#enable-redaction).
Option #1: Generate both redacted and original transcripts and audio.
Option #2: Generate redacted transcript only, and both redacted and original audio.
Based on the options in the redaction, it seems to me to suggest that "original audio" will always exist?
Down the way a bit when discussing using dynamic contact attributes we have the following options.
Option #1: None
Option #2: RedactedOnly
Option #3: RedactedAndOriginal
Hmmmm. "RedactedOnly" you say?
I went ahead and setup a dynamic attribute and passed this to the recording and analytics behavior block thinking that I would only get the redacted information after post-analysis has been completed. ie: Delete the original file after it's done it's job.
Am i missing something? Gotcha's not documented? Oversight on fundamentals?
AWS OFFICIALUpdated 4 months ago
Hello! Thank you. I'm not sure what happened originally when i tested where i thought deleting the original audio also deleted the redacted. I tested again and that was not the case, so this helps significantly.