Accuracy on speaker attribution is where automated transcripts typically fall apart, and that's a gap here. The AI chat works well for general content questions, but multi - speaker videos or heavy accents will likely produce less reliable transcripts, which undermines the entire value proposition for interview - heavy workflows.
Hey Jeroen! Thank you for the kind words and the feedback, this is super valuable! Im happy to hear you found it useful!
I recently added a feature to quick-download the transcript alone without waiting for AI summaries, thank you for the feedback!
I agree on the multispeaker point! Would research ways of covering this as well as heavy accents!