This woman entrepreneur is building an AI-powered search engine for fashion products

About four years ago, when Divya Manjari was looking online for a particular cushion cover, she was not satisfied with the results she obtained. Sifting through various websites to find the best product and price proved to be challenging for her. 

This experience led her to explore the idea of an ecommerce platform where users could easily explore various brands, compare prices, and check sizes. So, in 2020, Manjari started her venture, Redfynd, an AI-powered ecommerce search and discovery platform for fashion. 

Interestingly, though she was from a business family, Manjari had never wanted to start her own business. 

“I was always of the opinion that a job worked better for me. But my experience in various companies, particularly in the ecommerce space, led me onto the path of entrepreneurship,” says Manjari, who is now based in San Francisco Bay Area, 

After graduating from IIT Kharagpur in 2001, Manjari joined IBM in Bengaluru as a design engineer and worked in hardcore tech roles till 2008. 

In 2008, she began working as a supply chain and merchandise portfolio manager. In her role, she led new initiatives and improved the supply chain and merchandising for the company. Later, she also went on to work as a senior program manager at Infosys for retail and ecommerce.

At Infosys, Manjari helped build technology to enable business growth. She also led the effort to build a predictive tool that monitored and estimated job run time for thousands of backend jobs and created alerts for critical business impact.

As the ecommerce product manager at Unilever, Manjari led the global rollout of ecommerce sites for various brands. She also streamlined the technical landscape by implementing CI/CD (continuous integration and continuous delivery/deployment) processes, automated testing, and monitoring systems. She also went on to work as the senior product manager of Azure at Microsoft. 

Manjari believes her diverse work experiences gave her a better understanding of the technology and ecommerce industries. She also thinks her executive MBA in 2019 acted as a catalyst that pushed her onto the path of entrepreneurship.

During the Covid-19 pandemic, Manjari noticed that online shopping surged. And even after the pandemic, users continued to shop online due to its convenience. She also recognised the need for a platform where users could easily explore various brands and compare prices.

This realisation, sparked by her own experience shopping online, inspired her to start up on her own. 

Manjari believes that, while there is significant advancement in AI across fields, its implementation is lagging in ecommerce. 

“At Redfynd, AI plays a crucial role in enhancing the online shopping experience, personalising customer interactions, and making search intuitive,” she says.

Bengaluru-based Redfynd uses machine learning algorithms to analyse users’ browsing behaviour, preferences, and past purchases to deliver personalised product recommendations. It also takes into account the user activity around filters, searches, and category navigation. 

“For example, which sizes did they filter, what discount percentage, did they sort by price, discount or relevance, and which product category, colour or fabric type did they go for most. This information feeds into building the deals recommendation on the home page and ranking results on search pages,” explains Manjari. 

The platform also uses computer vision technology that allows users to upload images of fashion items they like. AI then analyses these images to find similar or identical products available on Redfynd’s database of items, from over 25 million images. 

“This feature leverages computer vision techniques to interpret elements like colour, shape, and pattern, which help users find products they desire with high accuracy. Redfynd finds the same product sold at different stores by comparing the images. So we can find it even when the brand name, title or background in the image is different,” says Manjari. 

AI is also used to extract detailed attributes from product descriptions, images, and specifications. 

For instance, from the product image, AI can identify and tag attributes such as colour, material, style, pattern, and size without manual intervention. Manjari explains that this capability helps in extracting and combining attributes that may not be listed by the online store. 

Redfynd will soon add a generative AI feature which will help users find products by chatting with a personal assistant. 

“Instead of searching and applying filters, users would be able to chat and ask the assistant to refine the search results–very similar to how they interact with an assistant in a store. You could search for a white dress and then ask it to show only printed and short dresses from the designer store,” says the entrepreneur. 

The platform has partnered with ecommerce websites such as Myntra, Amazon and various small brands.

Once the user selects a product, they are redirected to the brand’s website or an ecommerce site to make the purchase. Users earn cashback when they go to the partner websites from Redfynd and make any purchase, not just the item they clicked on. They also earn cashback for referrals. 

Ecommerce partners are charged a commission on sales, while brands are charged based on referral traffic.

Most loved story

---

(100 Emerging Women Leaders) This young entrepreneur is building a sustainable beauty brand with simplicity at its core

Challenges along the way

Manjari says her journey was far from easy, particularly when she started her venture during the lockdown while caring for two small children.

“It was all overwhelming. I had two kids, a family, and a startup to manage,” she says.

She acknowledges the pressure many women feel to juggle everything, both at home and at work.

“We tend to place a lot of burden on ourselves to manage it all,” she reflects.

However, she is grateful for her husband’s unwavering support during this challenging time.

“During Covid, when our helpers couldn’t come, my husband and I managed everything together,” she recalls.

“It was a roller coaster ride, and there were moments when I didn’t think I would make it. But I did.

Challenges will come, but you need to pull up yourself and keep moving ahead,” she says.