Description
I'm building Hula, an AI assistant that lives inside iMessage. Users just text it, and it can search across their connected apps like Gmail, Calendar, Slack, Notion, Asana, and Todoist, pull together an answer, and take real actions on their behalf with confirmation before anything happens. Most of the product is already built. Backend, mobile app, the integrations themselves, and a cross-app reasoning layer that lets Hula pull context from multiple apps at once. It's been built fast, with heavy AI-assisted development, and while a lot of it genuinely works, I've run into real problems getting it fully stable. Things get built and reported as working, then fail when actually tested live in the app. Some features work in one specific flow but not in others they should also cover. I'm being upfront about this because I want someone who can look at the real state of the code honestly, not someone I have to sell a polished story to. I have a full technical audit ready covering the entire codebase: what's confirmed working, what's confirmed broken, and what's never actually been properly tested against live behavior. I'll share this with anyone I get on a call with. What I need: Come in and independently verify what actually works vs what doesn't, using real, live testing, not just reading the code. Help me close the real, confirmed gaps fast. Be comfortable working inside a large, partially built codebase (190+ files currently uncommitted) rather than starting fresh. Be honest wi