Absolutely amazing, I've always wanted to do something with syncing transcription and timestamps but was looking for some guidance. Thank you incredibly for making it open source.

Not sure why this isn't getting liked more, but maybe you could try some of the cheesy cuts like https://www.youtube.com/watch?v=6zz-lTa1SAM

Either way, I'm sure the guys at comedy news shows will love this.

Could you take this a step further and automatically run the transcripts created through the Watson Personality Insights?

Oh yeah, there are a ton of possible combinations...with and within the Watson system. I haven't found much success with their currently available vision APIs, but have thought it'd be a cool idea to calculate facial expressions during given phrases, which is something that Microsoft Oxford and Faceplusplus offer...though that's probably not as relevant as calculating vocal tones.

good job