Shopify frames taxonomy at scale as three problems: volume, expertise, and consistency.

When you’re operating a taxonomy with 10,000+ categories, manual review will not work. and by the time you react, merchants are already listing products that don’t fit.

This piece breaks down how Shopify moved from reactive manual updates to a multi-agent system that scans taxonomy branches in parallel, proposes new categories/attributes from merchant data, detects duplicates via equivalence relationships and runs automated QA through domain-specific judges.