Building a conversational AI startup from zero to one

Local note

This was written before AI became the default investor and product buzzword. The useful part is not the hype around conversational interfaces. It is the practical question underneath: can software understand enough context to change how work gets done?

I keep it linked here because it is part of the same thread I still write about: software companies, AI, markets, and the work of turning an unclear idea into something people can use.