Overview of My Submission
Utilizing Deepgram as a virtual speech input in a web desktop.
Link to Code on GitHub
Additional Resources / Info
Video of the most useless demo ever: https://user-images.githubusercontent.com/1139657/162651792-32cb216c-2257-40f5-a75d-0b2062b65897.mov
This is a project that I actually take great pride in, however, when trying to balance development time with my full-time job as well as spending time with my family, I wasn't able to complete the project to my satisfaction, so I called it, "The Most Useless Demo Ever."
My Deepgram Use-Case
For several years some buddies and I have been talking about using virtual speech inputs in some programs that we are writing. And we've done that with varying degrees of success.
I've personally had a strong desire to make a web desktop over the past 10 years or so which is inspired by many concepts in Unix-based operating systems.
So I'm presenting the largely incomplete ReShell desktop system which features mobile and desktop paradigm support.
There are several variants of this desktop, some of which are dedicated for peer-to-peer (P2P) calling.
In time, these multiple variants will converge into a single "portal" and become decentralized, where users can bring their own backends.
This Deepgram submission is basically my way of saying, "we support voice, are supporting multiple input interfaces, and we have a long road ahead."
In our UI, I've integrated Deepgram and Azure's speech recognition services, and they can be run concurrently. If wanting more feedback as to the pros / cons of each and what I have experienced, feel free to reach out to me.
You can find me on LinkedIn at: https://www.linkedin.com/in/jeremyharrisconsultant/
Top comments (1)
Hahah. It's definitely not the most useless. There's a lot of interesting features you have in there!