Real work starts with a problem you're trying to solve, a project that's half-formed, and a wall full of sticky notes that haven't become tasks yet. It's messy before it's structured, and then you can't ideate and be creative within one space; you have to plan in one, draw in another, while referencing a pile of sticky notes and a whiteboard... it's a mess.
Chimerical starts differently. You type a description of what you're working on, however you think about it, in plain language. The AI classifies your intent (is this a creative project? a product build? a research sprint? home improvement? etc.) and constructs a workspace around your answer featuring the right tools, a named project, and a creative ideation space ready to use.
The more interesting thing is what happens after that onboarding.
That's the thing I'm most proud of, something I call Canvas Awareness. The canvas observes how you work: what you touch, what you leave behind, and how your thinking evolves. It speaks in one consistent voice, and it never directs you. Not a chatbot; it's more like a muse pointing things out. It just notices things - and occasionally, it says something that makes you realize the canvas has been understanding your project more deeply than you expected, getting you through a blocker or inspiring the next idea.
Under the hood, every canvas wipe deposits a "geological layer". The AI captures what you seemed to be working toward, whether you were making progress, and a rough summary of the ideas present. When your current canvas shows alignment with a stalled layer from weeks ago, the canvas surfaces a single observation. It just gets you!
Tech stack: React 19 / Vite SPA on Vercel, Supabase for auth and storage, Claude API for classification and awareness, PostHog for analytics, built via Claude Code using my 16 years of domain knowledge of PM tools.
Went from zero to launch in 47 days. Happy to answer questions about the technical decisions.