You do not need a finished specification. A problem, an idea, or an uncomfortable “we know this should work better” is enough to begin.
A fragile application, stressful deployment, manual hand-off, or AI idea that needs a reality check.
I’ll read the context, ask useful questions, and suggest a sensible next conversation or course of action.
The goal is to make something meaningfully simpler, safer, faster, or more dependable.