Gil PaulApplied AI for business Case study 02 / Shelfie
Case study 02 · Levinsohn Textiles

Shelfie. Automated shelf health assistant.

Shelfie, an AI agent that compares live listings on digital ecommerce platforms against your real product information and available inventory.

Every company feels pressure to launch new listings, while old ones are forgotten. Listings break. Platforms change. Product updates, new images and videos never make it to the page. Too often, a problem with a listing is found only when someone stumbles on it.

Shelfie compares four sources of information and suggests the action required. The sources are your current inventory, retailer portal reports, the live product page and your team's running record of listing changes and work in progress.

Shelfie flags items shown as out of stock or unavailable when your inventory says you can ship, item variations that appear broken, and missing carousel images. Shelfie calls out each potential problem with a direct link, so human verification is a one-click task.

How Shelfie works

Four sources · a person at the gate
The Shelfie workflow On a schedule you set, threads run from four sources: the team workbook, inventory, the retailer portal and the live product page. They merge and pass through reconcile, grade every listing and check twice. A person reviews the findings at an approval gate, then findings are shared with the team, sorted as healthy, false out of stock, overselling, went inactive, or needs a person. Decisions made at the gate are written back to the team workbook.
The rules are yours

Inventory does not mean placement.

Having a color or size in the warehouse does not mean it belongs on every digital shelf. You set the rules. If Shelfie sees a gap, it asks the question rather than presuming the answer.

I thought I fixed that.

When someone marks a flag as intentional, Shelfie remembers and will not raise it again. If a decision needs to be reversed, a person makes that call.

These are Shelfie's rules today, and I'm glad to work with you to customize any tool to fit your needs.

Gil Paul · gilpaul.com01 / 02
Gil PaulApplied AI for business Case study 02 / Shelfie
What it changed

A tedious task, too often set aside, is now done.

Checking older listings has always been a necessary evil. By hand it could be a full-time job, so it never seemed to get done. New product launches always came first.

Shelfie has made it manageable. The agent works in its own environment and brings key stakeholders only the listings that need their help.

If you sell online, let's customize Shelfie to do this for you.*

*How closely a person partners with the agent depends on the ecommerce platform. Some are welcoming, while others have become more hostile. Either way, Shelfie gets the job done even if some platforms require more partnership from a human handler.

Work with Gil

Let's find your Shelfie.

I bring 40 years of business experience and hands-on experience building AI tools that people use every day. We start with a conversation about the work your team finds most tedious. A defined need becomes a project proposal. A broader question can begin with a focused assessment.

We agree on deliverables, fees and timing before any work begins.

I created Shelfie at Levinsohn Textiles, and I present this case study with their permission.

Gil Paul · gilpaul.com02 / 02