"I'm fine" means something completely different based on how it's said, but your AI can't understand it. I began a PhD at Stanford two years ago because nobody knew how to make speech models that *actually* understand what people mean. I paused that PhD and started @OrukLabs because we finally figured it out. Now, I'm excited to share that we joined of a16z @speedrun to become the speech understanding layer for every human-{insert technology} interaction. (Oh, and we're also the first ever company to be selected for both alpha and speedrun.) Thank you to my cofounder @Viangren, our amazing team from @Stanford , @UCBerkeley , @Cambridge_Uni (we're hiring!), and the wonderful folks at a16z speedrun: @emilybenn12 @andrewchen @KatiaAmeri @justmazer @ndrewlee @Tocelot @joegarciaisme @filgran_ (among MANY others). We're in San Francisco through Demo Day. If you're building anything that listens to people, we should talk.