Answers on the Store Floor: An AI Assistant for a Global Denim Brand
17K+
Retail product records and 550+ store documents indexed.750+
Stores running the assistant on the floor.80%+
Search time cut, the brand reports.
About this engagement
- Industry
- Retail
- Users
- Store stylists and store executives
- Timeline
- About six months
- Services
- Enterprise AI Knowledge Systems
A global denim brand runs its store floor on documents: product details, styling guides and the processes a store follows every day. A stylist with a customer in front of them needed one place to ask, in plain words, and get the answer back before the moment passed.
Two people ask it different things. A stylist asks about fit, fabric and what goes with what, across the Men's, Women's and Kids' ranges. A store executive asks how a range is selling and what is left in stock. Both ask in plain words and both ask in the same place.
The assistant is live in stores today. It reads across 17K+ retail product records and 550+ store documents, and it shows where each answer came from, so the answer can be checked rather than taken on trust. SoluteLabs designed and built it.
The build ran about six months. UX interviews and a design sprint came first and decided the shape of the thing: what a stylist would ask, how little they should have to type, and what the assistant has to show back before an answer is worth using on the floor.
Solution
One assistant sits over the brand's own store documents. A stylist asks in plain words and gets the answer back in the same conversation, with the source it came from shown alongside it, so the answer can be checked instead of trusted.
It is built on Google Cloud. The floor decided the shape of it: one place to ask, an answer in the time a customer will actually wait, and no second system to open while they are standing there.
Grounded answers with the source shown
Built on Google Cloud
Deliverables
Where the question starts
- Suggested questions sit under the box, so the ones a stylist reaches for most are one tap away.
- The same assistant on the desktop in the back office, the tablet at the till and the phone in the aisle.
The assistant on the floor
- A stylist asks about fit, fabric and styling. A store executive asks how a range is selling and what is left in stock. Both ask in plain words, in the same conversation, with no second system to open.
- Send a photo when showing the thing is faster than describing it: a garment tag, a shelf label, a printed process sheet.
The knowledge pipeline
- 17K+ retail product records, across 6.3K+ product codes and 1.4K+ style families, indexed alongside 550+ documents on products, styling guides and store processes.
- Product class, fit, fabric, lifecycle and channel are all searchable, so a question can be asked the way a person would actually ask it.
- Updated content lands in the same index, so what the floor is told matches what was last published.
Outcomes
Three audiences, argued from what the assistant does. No platform metric was invented to fill a fourth.
For stylists: the answer arrives while the customer is still there
A stylist answers without leaving the floor to go and look something up.
New and seasonal hires give the same answer as someone who has worked the floor for years, from their first shift.
For store executives: the range and the stock, answered in the same place
Every store answers from the same documents, so guidance stops drifting from one location to the next.
How a range is selling and what is left in stock come back in plain words, against last year and the year before.
For customers: the same answer, whoever is helping
Questions on fit, availability and styling are answered in the aisle rather than after a trip to the back office.
The answer does not change depending on who happens to be on shift.
Talk to Us About Your Store Floor
If your people are searching for answers while a customer waits, the documents are usually fine and the way in is not. Let us look at yours.
