Skip to content

Instantly share code, notes, and snippets.

@rpfilomeno

rpfilomeno/Customer Identification by Voice print

Created October 23, 2015 14:11
Show Gist options
  • Save rpfilomeno/a29a56b4bf1fd5fc482d to your computer and use it in GitHub Desktop.
Save rpfilomeno/a29a56b4bf1fd5fc482d to your computer and use it in GitHub Desktop.
Download ZIP
Call Center Techs
Raw
Customer Identification by Voice print
Extract voice fingerprint from previous customer call recording and store it for future
comparisons when the customer calls again.
Application:
* Fraud detection - detect the person voice print is different from the identity provided
and raise flag for review or detect known fraudulent individuals in calls and alert the agent
to inquire for additional indentification challenges.
* Identification - provide agents with additional feedback for user's authenticity of
his/her claimed identity
Candidate techs:
http://echoprint.me/
Raw
Voice Stress Detection
Detect stress in voice recording as basis for service quality. The idea is if theres a lot of
stress then the call didnt went well and could be flagged for review.
Application
* Fraud detection - detect person is lying during calls
* Customer service quality - detect irritations during calls (shouts, etc)
Candidate techs:
http://liarliar.sourceforge.net/
http://cmusphinx.sourceforge.net/wiki/ - has asterisk integration
http://www.politepix.com/openears/ - for mobile
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment