India’s Myntra innovates with generative AI to assist buyers put the correct look collectively


Bengaluru, Karnataka, INDIA: For somebody pushing 40, stepping right into a fitness center for the primary time could be a nerve-wracking expertise.

As this author realized, even earlier than self-doubts crept in about their possibilities of surviving an hour on the fitness center, the bigger looming query was what to put on and never look misplaced.

With completely no concept what they had been on the lookout for, this author turned to an AI procuring assistant by Myntra, India’s largest on-line vogue retailer, and typed, “I’m on the lookout for garments I can put on to work out within the fitness center.”

Surprisingly, the AI assistant understood precisely what this author wanted and got here up with jerseys that might wick off sweat, compression t-shirts, self-proclaimed comfy trackpants that wouldn’t limit motion, footwear that might make you run higher, health bands and all types of substances a beginner couldn’t have imagined they needed or wanted.

With the procuring cart full and the pockets considerably empty, this author was prepared for a brand new starting.

What the AI assistant did – convert an summary person question into actionable outcomes – is sport altering for the style business. Typical search works finest with particular key phrases – a blue t-shirt from a specific model, say.

It goes a number of steps past standard search. It makes use of generative AI to reply to extra open-ended questions like what to put on for a specific competition or a cricket match and even the trending vogue in a metropolis.

“That is huge,” mentioned Arit Mondal, director of product administration at Myntra, “Why? As a result of, that is the primary time we’ve got an answer, which is fixing the unsolved ‘search’ drawback within the vogue, magnificence and way of life business. And it’s dwell for patrons at scale.”

Portrait photo of a man standing in front of mannequins
For Arit Mondal, director of product administration at Myntra, generative AI has helped them remedy one of many largest search issues within the on-line vogue retail business. Photograph by Selvaprakash Lakshmanan for Microsoft.

Because the starting of on-line vogue retail, trying to find merchandise has been very comparable to looking for some other piece of data on-line. You strive a set of key phrases and hold refining your search with completely different key phrases and preset filters. 

A seek for a branded, blue t-shirt works effectively as a result of the key phrases are already a part of the product catalog.

However that’s not all the time how individuals store in the actual world. Some buyers solely have a obscure concept what they need – as an example, garments for an upcoming trip or a rock live performance.

The traditional technique of looking out by key phrases fails spectacularly relating to the second form of buyer because the search strings they use usually are not retrievable instantly from the knowledge saved within the product catalog.

Till now.

When generative AI – constructed on massive language fashions (LLMs) that synthesize huge troves of information to generate, textual content, pictures and extra – first made information final yr, the group at Myntra shortly started occupied with how they might leverage it to boost buyer experiences.

When Myntra organized a hackathon in February this yr, a gaggle of engineers from the corporate’s search group determined to make use of Azure OpenAI Service to unravel the summary search drawback and unshackle customers from the cuffs of key phrases.    

Photo of a group of people looking at a computer
The hackathon group at Myntra constructed a working prototype of the AI procuring assistant over two days. Photograph by Selvaprakash Lakshmanan for Microsoft.

They had been pleasantly shocked to see how ChatGPT, the generative AI service out there by way of Azure OpenAI Service, may synthesize pure language prompts. They requested ChatGPT in regards to the look of an actor from a current film and it may inform it consisted of a bomber jacket, gloves and aviator sun shades.

“And that is the knowledge that Myntra’s present catalog didn’t have,” mentioned Swapnil Chaudhari, an engineering supervisor at Myntra. 

Over two days, his group took over a convention room and stored making an attempt new prompts – textual content that generative AI may perceive – to see what outcomes they received. This was new territory – they usually didn’t understand how far they might push.

“We had been shocked to see the outcomes. It was capable of reply questions like garments to put on for regional festivals like Pongal and Onam,” mentioned Pragna Kanchana, a frontend engineer at Myntra.

On a whim, she tried to go looking in Hindi with sardiyon ke kapde, which in English interprets into winter garments. And it understood it!

The group then received entry to Azure OpenAI Service’s playground that allow them do rather more than was potential with ChatGPT alone.

“Leveraging Azure OpenAI Service, we had been capable of plug in numerous massive language fashions in the identical immediate and determine which mannequin labored finest for our use case. So, we had a variety of freedom to match and select the correct mannequin,” defined Santanu Kanchada, a backend engineer within the search engineering group.

The group knew they had been on to one thing huge. They wrote the code in a day, and inside two days they’d a working prototype of a brand new characteristic that enabled customers to go looking with pure language.

“If it weren’t for GPT fashions, we’d need to first retrain the mannequin utilizing Myntra’s catalog after which wait and test the outcomes with our expectations. However the pre-trained fashions already out there with Azure OpenAI Service had been already performing fairly effectively,” added Chaudhari.

Over the subsequent 5 weeks, a number of groups throughout engineering and product growth fine-tuned each the backend and the person interface for the AI procuring assistant.

“Myntra’s programs are on Azure and deploying Azure OpenAI Service was as seamless as deploying one other server and it gave us a safe approach of utilizing generative AI,” defined Vindhya Priya Shanmugam, director of engineering at Myntra.

Portrait photo of a woman standing in front of a wall of oversized colorful spools of thread
Vindhya Priya Shanmugam, director of engineering at Myntra, discovered Azure OpenAI Service as a seamless and safe approach of deploying their AI procuring assistant. Photograph by Selvaprakash Lakshmanan for Microsoft.

Submit the hackathon, the search engineering group stored refining the prompts to get helpful outcomes for customers. One of many issues, as an example, was how to make sure that the response to a person’s question resulted in garments for under the gender the person is on the lookout for.

Within the weeks resulting in the launch, they educated the system on Myntra’s catalog and added guardrails so the outcomes had been restricted to the catalog.

The AI procuring assistant was launched on the Myntra app in late Might, simply in time for considered one of their largest marquee occasions, Finish of Cause Sale (EORS). It included pattern prompts that gave customers an concept of how they might use conversational language quite than key phrases.         

Since then, Myntra has already seen search queries broaden, providing new alternatives for product discovery. As an illustration, when somebody searches for garments they will put on to a seaside, not solely seaside put on but additionally equipment like hats, sun shades and footwear pop-up.

It has been phenomenal for Myntra.

“Customers who store utilizing the AI procuring assistant are 3 times extra more likely to find yourself making a purchase order,” mentioned Mondal. “As a result of it additionally helps customers uncover an entire look from a number of classes of merchandise, we’re seeing that on common they add merchandise from 16 p.c extra classes than normal.”

Whereas this author’s health transformation journey remains to be questionable, a number of groups at Myntra are already constructing new options based mostly on generative AI.

One among them will permit customers to decide on completely different classes of merchandise – tops, bottoms and equipment, for instance – and see how they give the impression of being collectively in an outfit. Myntra plans to additional improve it by introducing voice search and supply personalised outcomes. They’re additionally taking a look at how they will use generative AI to assist the shopper assist groups.       

Prime picture: Myntra’s AI procuring assistant powered by Azure OpenAI Service lets buyers uncover an entire look utilizing pure language prompts that may embrace locations, festivals, or different events. Photograph by Selvaprakash Lakshmanan for Microsoft.