Vocinity Creates Conversational Bots with Deepgram

TRANSCRIPT

Vocinity

Nathan Stratton, Founder and Chief Technology Officer

My name is Nathan Stratton. I’m the founder and chief technology officer of Vocinity,, a conversational voice and video application startup. Our technology stack is based on best of breed and we look for vendors that excel in a lot of different areas. And Deepgram was brought into that mix initially at an exploratory level about a year ago. And then over the last year, we found that they really helped us with latency as a primary factor and also with accuracy. 

If you look specifically at our use case, it’s real-time voice communications with video avatars. And the time between when the user finishes speaking and our bot is able to start speaking is something that we pay very, very close attention to. And with previous ASR vendors that we worked with, that latency was up to five hundred to all the way up to one thousand or twelve hundred milliseconds longer than with Deepgram.

With previous ASR vendors…latency was up to 500 to all the way up to 1,000 or 1,200 milliseconds longer than with Deepgram.

Nathan Stratton
CTO, Vocinity

Well, some of the best things about Deepgram, I think really start with the team. We’ve had a really good kind of level of discussion and working relationship with Deepgram in our unique use cases as well as some technical hurdles that we faced early on in the implementation. 

Being able to work with an organization that is as technically focused as Deepgram was very very beneficial. And I wanted to kind of refine that kind of relationship, you like to share that with others. Then also being able to find the accuracy with background noise was particularly helpful for us. And there were a couple other companies that we work with in the restaurant food ordering business that we thought could benefit from that technology as well.