Text-to-action models

At Adept, we’re training a giant neural network to do anything a human can do on a computer. It's a new category of foundation models (text-to-action) that can use every software tool, every website, and every desktop app. In the long-term, we want Adept's natural language interface to be the easiest way to do most things on a computer (or phone, tablet, etc.), shifting the way we interact with computers from mouse / keyboard / GUI to natural language.



Founded 2021



San Francisco, California