This is a simple demo of the Microsoft Cognitive Services Speaker Recognition APIs:
By using your browser's built-in audio recording capabilities, and a custom version of the fantastic RecorderJS to convert the raw audio into the format required by Cognitive Services, you can easily have a play around with using your voice as a form of verification and also to identify who is talking out of a group of people.
Super cool stuff.
You can view the demo itself here:
Articles that support this are over here:
Also you can view the LDN BotFramework Meetup session I gave using this over on YouTube: