PupManager started from a real problem: dog training businesses running bookings, classes, client records and payments across spreadsheets and paper. I set out to design and build a proper multi-tenant SaaS product for it — solo — using AI-assisted workflows.
What AI made possible
The biggest shift was how quickly I could move between roles. As a designer I'd map a flow; as a builder I'd have Claude help me reason about the data model and implementation; then back to design to adjust. The loop was tight.
- Explored product structure and edge cases early with Claude
- Drafted flows, logic and technical approaches fast
- Moved from problem to a working, multi-tenant product quickly
- Kept customer understanding and UX judgement in the driver's seat
What didn't change
The product still had to be right. A drag-and-drop scheduler that matches how trainers think. Registrations and payments modelled around classes, not generic ‘products’. An admin surface simple enough for non-technical owners. AI helped me build it — it didn't decide what to build.
Senior product thinking plus AI can take an idea to real, buildable software remarkably fast.
That's the model I'm most excited about: designers who understand the product deeply, using AI to ship far more than they used to.
← Back to all posts.webp)
