Nordstrom Product Information Management: Entity Detection

As a lead member of the UX team at Nordstrom, I designed a solution to eliminate duplicate products within our inventory systems .

The problem

Our product information management system (PIM) could contain duplicate records when we ingest new item information or when changes are made to item information if we do not have rules and logic to check for and merge existing products, colors, and items.

If the rules and logic are not able to determine if a potential match is a true match, we need a human user to review the potential/fuzzy match to make a decision.

 

Researching Current Pain Points

• Process requires uploading a file

  • Files can be large and it can be challenging to get through an entire batch in one sitting

  • Scanning all of the information and rows to find discrepancies can be challenging and takes some practice

  • Hard to keep track of which potential matches have been addressed and which are still pending

The Solution:

We planned to surface these potential matches in an easy-to-use user experience so that we can efficiently reduce the number of duplicate product records and uphold the integrity of our product information.

How we arrived at the solution

  • Defined current process - oiefihglakhgla

  • Collected requirements to complete jobs as they exist today - ehglkahgadg

  • Worked through high level flows to ensure we aligned - etc etc etc

  • Fleshed out designs to match our current Retail Hub platform visuals

 
 
Summary

This project required navigating a complex web of legacy processes—including email-based excel tracking and information scattered across disparate systems—and transforming them into a streamlined, AI-powered solution. I collaborated closely with my Product Manager to facilitate a 3-day workshop that prioritized user needs and established a phased roadmap, giving our team clear direction to successfully deliver the first release. The resulting interface was elegantly simple: users could quickly review AI-matched product data and make confident merge or create decisions when human judgment was needed. The impact was significant—users went from juggling over 10 research tools down to a single, intuitive interface, dramatically reducing the time spent on product consolidation while improving data accuracy and reducing inventory variances across Nordstrom's ecosystem. The combination of tackling organizational complexity while delivering a clean user experience that saved hours of manual work made this particularly rewarding.

 
Back to portfolio