Inside Myntra's AI playbook: Faster seller onboarding, smarter shopping, leaner operations

Myntra has reduced the time it takes for a new seller to go live on its platform from 10–15 days to about one or two days, as the fashion ecommerce company expands the use of artificial intelligence across its marketplace.

The announcement came at mynnovAIte, Myntra's technology showcase held at its Bengaluru headquarters on July 16. Led by Chief Product Officer Lakshmi Narayan Swaminathan and Chief Technology Officer Pramod Addidam, the event focused on how AI is being deployed across three areas: sellers, customers, and internal operations.

The underlying message was straightforward. AI isn't being used to solve new problems, Swaminathan said, but to solve familiar ones faster and more efficiently.

Helping sellers get to market faster

For Myntra, onboarding is only the beginning of a seller's journey. That journey includes listing products, securing a first order, and reaching what the company calls “activation”, the point at which a seller crosses 100 orders and unlocks additional growth opportunities.

The biggest change has come at the onboarding stage. Registration now takes only a few minutes, while complete onboarding is finished in approximately one to two days instead of the earlier 10–15 days.

Much of that improvement comes from automating catalogue creation.

Instead of manually entering product information for every listing, sellers can now upload a single product image, after which AI extracts structural attributes such as colour, fabric, neckline, and sleeve style. Myntra says the system identifies up to 40 attributes for each product, with catalogue experts reviewing the output instead of creating it from scratch.

The company is also tackling one of the biggest hurdles for smaller brands: product photography.

During a live demonstration, Myntra showed how sellers can upload a flat-lay image of a garment, choose a model, select a mood such as cheerful or confident, and specify a styling direction such as contemporary or elegant, complete with matching accessories. The system then generates catalogue-quality images from those inputs.

According to the company, what previously took about a day can now be completed in four hours. Myntra also says it currently generates between 400 and 600 AI-powered product videos every day.

The company has also redesigned its Partner Portal. Rather than presenting sellers with dashboards full of data, the platform now prioritizes ranked daily actions, from correcting listing issues and updating prices to improving visibility through advertising and growth tools.

Myntra has also introduced Saarthi, a voice intelligence platform that calls sellers when operational issues or pricing gaps need attention. Built as a configurable platform, Saarthi can also be adapted by other internal teams to create their own voice-based workflows.

Swaminathan said seller feedback shaped many of these changes. Sellers wanted simpler ways to resolve incomplete or stuck listings and clearer guidance on what actions to take next instead of interpreting large volumes of performance data.

“There is more trust in the recommendation that we give when a seller is told exactly what to do next rather than handed a dashboard to interpret alone,” he said. 

On whether these AI tools would eventually become paid features, Swaminathan said Myntra currently has no plans to differentiate access based on seller tiers, adding that the tools are intended to improve performance rather than serve as premium offerings.

Fashion trends also change rapidly, making it difficult for sellers to keep pace. Swaminathan said part of Myntra's role is to translate that speed into recommendations sellers can act on.

Bringing the trial room online

On the customer side, Myntra is using AI to reduce one of online fashion's biggest challenges: uncertainty around fit.

Its size recommendation engine now covers 85% of the eligible apparel catalogue and delivers recommendations in under two seconds. A fit visualizer, which is still being scaled, shows how garments are likely to drape on a body type similar to a shopper's, while an explainability layer tells users why a particular size has been recommended instead of simply displaying one.

The company has also introduced a styling layer that allows shoppers to mix and match outfits on a single model. Another feature, ‘Try it on Me’, currently being rolled out, is designed to show how the same outfit could appear on the shopper’s own photograph.

According to Myntra, more than 90% of its monthly active users already receive personalized search results.

Its AI-powered conversational assistant, Maya, has recorded a 25% repeat interaction rate among engaged users. Myntra also said that 40% of engaged users have become repeat users of its AI-powered features.

The company's social commerce feed, built around a creator community of more than six million registered shoppers, is now used by 25% of its monthly active users. Sessions per user have increased by 8% to 10% since launch, although Swaminathan noted that the affiliate programme itself is only about a month old.

AI behind the scenes

Beyond customer-facing features, Myntra is also applying AI across its internal operations.

Addidam said the company's focus is shifting from using AI primarily as a question-and-answer tool to using it as a decision-making platform.

One example is BIRA, Myntra's internal natural-language analytics platform. According to the company, it reduces analysis that once took days to just minutes, delivering insights around 10X faster.

Another is Meera, the company's customer support assistant, which now resolves 31% of customer queries. Myntra says satisfaction scores for those interactions have doubled.

Across engineering, the company says AI has increased feature rollout speed by 40%. It has also reduced supply chain network simulations that previously took two days to about an hour.

Looking ahead, Myntra is building a seller-facing copilot designed to help account managers support sellers more effectively and consistently. Swaminathan described it as “an equivalent of ChatGPT”, though the tool is still under development.

Taken together, the announcements point to a broader shift in how Myntra is approaching AI. Rather than introducing standalone features, the company is embedding AI across the seller journey, customer experience, and internal operations, with the aim of removing friction at every stage of its ecommerce ecosystem.