Overview

Catalog AI is a natural-language design discovery system built for the jewelry manufacturing industry. It enables customer-facing teams to search an existing design catalog using structured, natural-language requirements instead of relying on a multi-stage internal process to curate and share matching designs.

The challenge

When a customer requests designs matching specific criteria, such as tennis bracelets under 5 grams of gold with a defined diamond weight, the request traditionally passes through customer representatives, department heads, and product development teams before a curated selection reaches the customer. This creates delays, introduces opportunities for information distortion, and makes repeated rounds of customer feedback expensive in terms of time and coordination.

Approach

I developed a natural-language database search workflow that translates structured product requirements into a search for matching jewelry designs. Users specify relevant attributes, such as category, subcategory, gold weight, diamond weight, and color-stone weight, using a familiar query format. They can review the retrieved designs, select the relevant pieces, and export a PDF catalog to share directly with the customer. An image-based search capability is also under development.

Key features

• Natural-language search across the jewelry design database
• Structured product requirements covering categories, subcategories, and material weights
• Retrieval of matching jewelry designs from the existing catalog
• Design selection for customer-specific requirements
• PDF catalog generation and export
• Image-based jewelry search under active development

Takeaway

Catalog AI addresses a workflow problem at the intersection of AI and jewelry manufacturing operations. By giving customer-facing teams a more direct way to discover and share relevant designs, it reduces dependence on repeated internal handoffs and creates a foundation for a faster, more responsive customer experience.