Nicole Ruiz, a writer and parent, has built a Claude-powered purchasing system that filters out drop-shipping operations and surfaces century-old manufacturers over algorithm-optimized direct-to-consumer brands. The core is a Claude Project with custom instructions that evaluate brands on heritage, craftsmanship, and return policies. The practical demo in this interview is a can opener search, starting at the 10:30 mark, where the system navigates vendors like Boston General Store and Manufactum despite their notoriously poor UX.
The argument Ruiz makes that earns its own attention: AI levels the playing field for small artisans and heritage brands that cannot compete with Amazon's infrastructure but make objectively better products. Claude Cowork handles the return side of the equation by pulling receipts from email and drafting refund requests automatically. The drop-shipping vetting technique, covered at 26:33, gives you a repeatable filter for separating legitimate manufacturers from repackaged commodity goods.
The full interview covers seven distinct workflows, including budget-constrained shopping with gift cards and an 'anti-to-do list' framework for reducing household mental overhead. The methodology is transferable. Read it for the specific prompting architecture inside the Claude Project, not just the shopping philosophy.
[READ ORIGINAL →]