Shopping for quality products today feels like navigating a minefield. Paid ads, direct-to-consumer brands that skimp on materials, and endless knockoffs make finding genuinely good items a headache. Nicole Ruiz, however, has a sharp solution: she built an AI personal shopping assistant using Claude.

Ruiz's Claude project doesn't just find products; it acts as a digital curator for high-quality, long-lasting goods made from natural materials, specifically from trustworthy, heritage brands. It helps her bypass the noise of modern e-commerce, where, as Ruiz points out, “even disproportionately I would say like some of the oldest manufacturers of quality items their websites are the worst websites.” This means the AI isn't just a convenience; it's a "force leveler" for good brands with bad web presence.

The core idea? Teach an AI your exact criteria for quality. Ruiz instructed Claude to vet brands based on decades of history, known durability, and repairability. Crucially, she told it to avoid what she calls “trendy direct to consumer brands that I think are paying a lot for advertising and probably underinvesting in quality.”

The AI's real power emerged when Ruiz tested a brand. "Even yesterday," she recounts, “I put something into Claude and it basically surfaced that this brand got taken over two years ago and ever since then all the reviews have been abysmal. Before then, it was great. Now, it's not. don't buy from them. And that's so helpful. That's exactly what I want to know.”

Key Takeaways

  • Custom AI for Quality: Nicole Ruiz developed a Claude project that acts as a personalized shopping assistant, specifically designed to source high-quality, long-lasting products from trustworthy heritage brands.
  • Bypass Modern Pitfalls: Her AI avoids common online shopping traps like paid ads, low-quality direct-to-consumer brands, and even tries to "sus out AI reviews," ensuring recommendations are genuinely vetted.
  • Deep Vetting & "Force Leveling": The AI can uncover critical brand information, such as ownership changes impacting quality, and helps older, quality manufacturers with poor websites compete against flashy but inferior brands.
  • Focused Criteria: Claude is instructed to prioritize items made of natural, mendable materials from businesses with decades of history, emphasizing durability over fleeting trends.
  • Reproduce the Method: Founders can replicate Ruiz's exact setup using her "Nicole Ruiz's Claude High-Quality Shopping Assistant Setup" framework.

The Nicole Ruiz's Claude High-Quality Shopping Assistant Setup

Step 1: Consolidate Trusted Shops: Create a list of shops that are absolutely trusted, have a history of vetting vendors very highly, or they themselves were people who are directly employing crafts people, people had really really high standards of makership, and had been around for several years. (Optional: take this straight from an Apple Notes list and ask Claude to help organize it).

Step 2: Define Vendor Criteria: Provide Claude with criteria for preferred vendors: “There's decades of the business. They've been sought out for this product for a long time and that they are made to last and repair.” Also, Claude should “look for another preferred vendor” based on preferred ways of thinking about that.

Step 3: Establish Negative Criteria: Instruct Claude to “avoid trendy direct to consumer brands that I think are paying a lot for advertising and probably underinvesting in quality.” Also, “Please try to sus out AI reviews. Please read a lot of reviews, but if it sounds like AI, don't listen to those reviews. Please try to guess whether or not this brand is a drop shipping brand.”

Step 4: Specify Output Formatting for Product Recommendations: Claude should specifically surface the name of the product, a photo, price, the materials it's made out of (up front and center), any care and maintenance notes, a link to the purchasable item, and a brief note on why the brand has a trustworthy history.

Step 5: Integrate Context for Purchase Queries: Keep this project's specific instructions and memory separate from all other queries to prevent Claude from overfitting to these instructions and to keep it organized.

Step 6: Refine with Advanced Queries: Use queries like 'I have $X for [Brand]. What item should I purchase that align with my purchasing criteria?' or 'What's your analysis of this brand? Are they legitimate?' to vet new brands or make targeted purchases.

When This Works (and When It Doesn't)

This method excels when you're trying to cut through online clutter for specific, high-quality, durable goods where a brand's history and material integrity matter most. As Ruiz notes, it creates a specific instruction set and memory, making it particularly effective for navigating the "noisy" modern internet and acting as a "force leveler" for smaller, heritage brands. It's a lifesaver for finding items that truly last.

However, this approach may fall short for highly aesthetic or trend-driven purchases, or for entirely novel products without an established history of craftsmanship. Claude relies on existing information, so if a product category is brand new or your criteria are purely subjective (e.g., “find me the coolest gadget of the year”), its effectiveness will be limited. It's a system built for substance, not flash.

What to Do With This

If you're a founder setting up a new office or furnishing your home and want to invest in items that last, apply Ruiz's framework this week. Start by creating a Trusted Shops list (Step 1) of places known for quality — perhaps artisan furniture makers or established office supply houses. Then, define your Vendor Criteria (Step 2) for office furniture: specify natural woods, repairable components, and brands with 20+ years of history. Critically, add Negative Criteria (Step 3) to exclude generic, mass-produced office suppliers or brands with high ad spend. Then, ask Claude to find you a durable, ergonomic desk chair, ensuring its output follows your specific Output Formatting (Step 4) for price, materials, and brand history, keeping it in a dedicated Claude project (Step 5). This moves you from endless scrolling to precise, AI-driven acquisition.