Get App
you are here: HomeNewsTechnology
Last Updated : Oct 10, 2018 05:33 PM IST | Source:

IBM working on leveraging AI and cognitive technology for retail industry

Imagine that in a retail store, outside the fitting room, there is a cognitive mirror that analyses what you have worn and gives you suggestions

Neha Alawadhi @alnehaa

In a conversation with Moneycontrol, Sriram Raghavan, Vice President, and Amith Singhee, Manager, Retail and Operations at IBM Research India, discuss the work being done by the tech giant in India in the field of cognitive fashion using technology such as AI, natural language processing and computer vision.

Edited excerpts:

Q: Why did IBM get into research in the retail space?


Sriram: For good reason, there is a lot of attention on online e-commerce. Often what is forgotten is the significant and actual percentage of our retail economy that runs through brick and mortar retail.

One observation was just as technology has been used, both to dramatically alter user experience, and optimise and create this online shopping experience with e-commerce, how do these same technology building  blocks like AI (artificial intelligence) and blockchain help in a retail brick and mortar scenario?

We looked at how can the brick-and-mortar retailers use tech like AI to leverage this rich set of data that they have including what has sold, where it has sold, their current decisions on inventory and stock management and how do they use digital technology and take it to the net level, very much like he online players.

Many of us feel that when it comes to an understanding of my past purchase behaviour, online is often better than even in the physical space because somehow online world seems to know me better than the physical world.

Q: But won't greater personalisation in the offline world make the experience annoying if a salesperson was constantly asking you to try new things?

Amith: As a researcher, the interesting thing is that I am also a consumer.

So you want to fill gaps that exist today rather than creating new types of problems for customers. Lets say a fashion conscious young woman walks in to a store and she is doing some exploratory shopping.

Instead of a salesperson follow you everywhere, lets say outside the fitting room there is a cognitive mirror, and it analyses what you have worn, and it gives you suggestions to complete the look.

Maybe it recognises that you're wearing something with stripes and a boat neck, and it makes more suggestions to complete the look, suggests a size etc. That's more helpful that a person hanging around

For that you need some pretty serious AI that can understand looks and choice.

Sriram- That interactive AI system can tell you what you're wearing

Going into physical stores in India, movement is often hard because you don't know what is kept where. So this system can help you find that- you do all of that digitally, see what you want is kept where.

Q: So do you draw on retail experiences from markets elsewhere, like Singapore? How much of retail research is derived from a developed market?

Amith: I think its a pretty healthy mix of both.There is a lot of overlap because shopping doesn't look too different.

But there are some specific Indian problems.

Fore example, if you are a gold jeweller, lets say you have 100 stores across the country. You'll have some limited inventory within a store.

Lets say there's a Bengali woman living in Bengaluru, and she goes to a Malleshwaram (Bengaluru locality) jewellery store. The stock over there will be tuned for local sensibiolities and she might not find things that would help her in the wedding coming up in her community in Kolkata. So that's a lost sale for the store.

However, the jeweller does have that inventory in Calcutta or somewhere else. That's where we come in to say how do we encourage the person to try the stuff in the Kolkata store virtually using AI, image processing, augmented reality, etc. The customer can then choose and get it next day through the jeweller's logistics system.

Q: There is also a lot of focus on omni-channel in retail. A lot of  retailers want to build an online presence. How do you balance that? 

Amith- Much of this tech can be reappllied to boost online presence because it is software and AI driven.

Sriram: A lot of retailers are also looking to offer in-store fulfilment. A customer can buy online but if you want it right way, go pick it up in the store. The same stocking decisions in the store continue to matter.

 Q: How may retailers are you working with in India and what kind of work are you doing with them?

Sriram: Our three main segments have been fashion including apparel and jewellery. Now we're starting to get into food and grocery. Across these two, we are actively working with 4 or 5 retailers in food and in active conversations with a dozen in grocery.

Early pilots have been completed with two or three retailers, and there are active pilots going on with a couple more. Conversations on refining these are going on with a dozen more.

As we look at AI on the customer side and decision making side, we will see serious optimisation of the supply chain. What is very unique to India is that you have to do all this at a price point that makes sense to the customer.

The next wave of innovation in India will come from technologies like blockchain  that will likely come in to remove more and more friction from the physical supply chain.

Get access to India's fastest growing financial subscriptions service Moneycontrol Pro for as little as Rs 599 for first year. Use the code "GETPRO". Moneycontrol Pro offers you all the information you need for wealth creation including actionable investment ideas, independent research and insights & analysis For more information, check out the Moneycontrol website or mobile app.
First Published on Oct 9, 2018 09:14 pm
Follow us on
Available On