How Stitch Fix Is Using Generative AI To Help Us Dress Better

Enterprise leaders trying to harness generative AI – the know-how made well-known by ChatGPT – are going through some main strategic questions. What are the very best methods to make use of generative AI inside the group? The place can the know-how add probably the most worth? And how will you discover a delicate steadiness between human experience and machine intelligence?

On-line private styling service Sew Repair is a enterprise that each one organizations can study from. The corporate was constructed on knowledge and AI from the bottom up, permitting it to supply a customized styling service and ship garments on to clients’ doorways – garments that they know shoppers will love based mostly on their fashion preferences and former selections. However what makes Sew Repair actually particular is its capability to strike that good steadiness between human information and machine capabilities. Right here’s how Sew Repair is harnessing generative AI to streamline processes and supply a greater service to clients.

A Knowledge Goldmine

Sew Repair is sitting on a goldmine of practically 4.5 billion text data points that shoppers have shared with the corporate. This knowledge might embody a consumer’s fashion preferences or suggestions on how an merchandise matches them, for example. It is helpful knowledge for understanding every buyer’s fashion, but it surely’s usually within the type of freeform textual content – making it difficult knowledge to work with. Generative AI helps Sew Repair leverage this advanced, messy knowledge as a result of generative AI excels at shortly making sense of and summarizing huge quantities of textual content knowledge.

Sew Repair makes use of massive language fashions from OpenAI, mixed with their very own deep studying advice algorithms, to interpret the suggestions that shoppers share after which use that info to tell future suggestions. In different phrases, generative AI interprets buyer suggestions right into a format that Sew Repair advice algorithms can perceive extra simply.

Let’s discover some particular use instances in slightly extra element.

Creating Clothes Suggestions

One of many methods Sew Repair makes use of generative AI is to floor a choice of clothes gadgets based mostly on suggestions the consumer has shared about their fashion preferences and match. These suggestions are shared with a human stylist, who will then curate the ultimate items for the person consumer.

Contemplating Sew Repair has hundreds of items in its stock, this preliminary advice saves stylists helpful time and permits them to spend extra time specializing in the consumer’s private fashion wants.

Supporting Textual content Era

Generative AI isn’t nearly deciphering textual content. It may additionally create textual content (and pictures, and video, and extra). Sew Repair makes use of this capability to provide partaking advert copy that communicates the advantages of signing up for Sew Repair’s service. All copy generated by AI is high quality checked by human copywriters (with a 77 percent pass rate on the time of writing).

Generative AI additionally helps Sew Repair create informative product descriptions that give shoppers particulars about an merchandise’s options, specs, and advantages. Writing descriptions for hundreds of things is, as you possibly can think about, a time-consuming process. However having skilled OpenAI’s GPT-3 on earlier product descriptions (written by copywriters), plus suggestions from shoppers, the mannequin can now shortly create descriptions for copywriters and merchandisers to evaluation. Sew Repair says the mannequin can generate 10,000 product descriptions every 30 minutes, with every description solely needing a fast one-minute evaluation by an skilled earlier than approval.

Routinely Styling Outfits

One other fascinating use is Sew Repair’s Outfit Creation Mannequin (OCM), which makes use of generative AI to generate millions of new outfit combinations per day. OCM, which was skilled on tens of millions of outfits created by stylists, picks gadgets from Sew Repair’s present stock, in addition to from clients’ earlier purchases, to compile personalised outfit recommendations for shoppers. This exhibits shoppers find out how to fashion gadgets they could be taken with buying (or already personal) with different gadgets out there from Sew Repair. That is achieved by way of personalised emails and adverts and thru the client’s personalised procuring feed.

Key Takeaways

Briefly, Sew Repair makes use of generative AI to higher perceive what clients need – and, importantly, to strike a steadiness between automation and human expertise. By letting AI methods automate the time-consuming, repetitive duties, Sew Repair frees up human consultants to concentrate on their uniquely human capabilities like creativity, judgment, and constructing superior relationships with shoppers. Because of this I believe all organizations can study from Sew Repair’s strategy.