| I'd like to give your take on what might be a drop-in component for probably the most important part of the workflows for the past while. | |
| What I'm beginning to work on now, which is using this app called Voice Notes, which does transcription through, I think it's Whisper probably. | |
| You record with your microphone through a browser or through an app. It sends it off for speech to text transcription and then you get your transcription. | |
| It does webhook delivery as well. So that's the initial way that I'm getting a lot of information in for workflows ideas to do notes. | |
| And I have a whole plan for what I can do with the notes on the other side of the chain. | |
| Let's say that I wanted to stand up a self-hosted tool, a component, which only had the task of capturing the recording, capturing the stream from the user, sending it off for transcription and saying, and so on. | |
| So if I'm getting transcription sent, that goes off to whatever the transcription tools that we're going to use is, let's say OpenAI, and then it's going to deliver when it gets it back from the speech-to-text model, it's going to send off a webhook and from there I can do what I'm doing currently. | |
| So the question is, does it make sense to build that from scratch, a little lightweight voice capture interface, or does it make sense, or is there a component that would be very much primed and ready to do this? |