kanemarais

Interesting. I often get meeting transcripts with muddled speaker attribution. Have you tested how it compares to Granola/Gemini etc?

I genuinely need this. I use different tools for different things and it would be amazing if they all shared a memory layer. Not sure why the four source options don’t include a way to access the LLMs and tools themselves?

Banging design - perhaps one or two things I’d tone down a bit (numbers and dial). But overall nice work. Must have been fun to engineer.

This is very interesting. Ideal for high school level education. Don’t be surprised if most people don’t quite get it at first - it’s a new capability. Must have been a challenging build. Congrats on shipping it.

My fav video gen tool by far