Hello everyone, I'm Evren, a full stack developer based in the Netherlands. I enjoy building things and reducing friction where it makes sense. I discovered TalkTastic while trying to build something like it for my own needs.
Evren Özkan Hey, that's awesome you tried building something similar already! You'll definitely appreciate what we're doing here then. I'm curious - were you working on a local Whisper implementation, or were you also thinking about the screenshots and context aspects?
I've been experimenting with Aiko using a shortcut that runs a post-processing script on the transcripts. For me, pasting the rewrite to text field and copying original transcribe to clipboard was a good enough middle ground. I was planning to add another step that would capture the active window using quickgrab, process it through macOS's OCR, and use that as contextual information to help correct any transcription errors or typos, especially for technical terms.