Welcome And Guest Introduction

SPEAKER_01 0:10

Welcome to What's in the Box, the brand new podcast brought to you by Box Technologies, powering retail with purpose. Box Tech delivers innovative and market-leading customer engagement solutions that turn business ideas into a performing reality. From design and integration to ongoing support and maintenance, we're with you every step of the way. Now, this podcast is a little different to what you might be used to. It's audio only. I say totally unscripted, we kind of know what we're going to talk about. And they're usually around about 20 minutes long because we're busy, our guests certainly are, and we're sure that you are too. So let's get right into it. My guest today is a distinguished expert in the field of artificial intelligence with a solid academic foundation, holding a PhD in computer science. Currently, he serves as Chief AI Officer at Kingfisher, where his leadership contributes to the advancement of artificial intelligence and machine learning technologies within the organization. With over 15 years of experience as an accomplished data scientist and visionary leader, he has consistently harnessed the power of data to drive strategic decision making. Prior to his role at Kingfisher, he held pivotal positions in leading technology and retail organizations, including the Hutt Group and Shop Direct. In these roles, he spearheaded cross-functional teams to implement cutting-edge machine learning and AI solutions, resulting in measurable and transformative outcomes. I'm delighted to welcome Mosen Kazimpur to What's in the Box. Mosen, thank you very much for joining us.

SPEAKER_00 1:50

Yeah, thank you for having me.

Hype vs Reality In Retail AI

SPEAKER_01 1:52

Well, a clue was in what about what we're going to discuss was in the introduction there. It's AI. And you and I first met a few months ago at an AI in retail conference. So and of course, we've we've just uh last week, well this this will be going out a few weeks after NRF, so NRF was held a second week in in January, and a lot was talked about about AI at NRF. Somebody told me that around about half of all the sessions were on that topic. So let's just kind of kick it off with a with a question for you. Are we are we in an AI bubble or or are we all just caught up in the hype of it? What are your thoughts?

SPEAKER_00 2:35

Yeah, I think uh that's a that's a good question. But I but I think it's very important to distinguish hype and bubble. So of course, I think there is a hype around AI, a lot of hype, and I don't think we are helping with around that hype that technically make it more amazing than it is. I mean, there's no doubt that AI technology was the most amazing piece of technology available to us. But clearly there are some hype around that. But when we're talking about bubble, is about if it's real and if the impact that we're talking about is real, and at least at Kingfisher and Across our brand, we have seen some tangible benefit and real impact to our post-customer and our internal operation. So I don't, I don't, I wouldn't say we are in a bubble, but I 100% agree there is a huge hype around the technology.

SPEAKER_01 3:24

But it and one of the themes that I picked up on and be interesting to get your thoughts on it, uh just briefly, still on NRF, was that people, the narrative seemed to be that we're moving away from hype to now more to reality and real-world use cases. Now we're gonna come on a little bit later to what you're doing at Kingfisher and with the brand, such as B ⁇ Q and Screwfix and so forth. But is that what you're excuse me? Is that what you're seeing? That uh we're we're kind of now in the process of leaving the hype behind and really starting to get real with this.

SPEAKER_00 4:00

Yeah yeah, I wouldn't say leave behind completely. There are gonna be always a new piece of AI technology which comes with a new uh type of hype around it. But yeah, definitely more people realize the actual value. So of course, many people start talking about POC in the past. So we see all of or most of those POC now start transitioning to the production and see real real impact. And as you mentioned, you know, there are some new advanced developments that try to get AI into different ways of doing things, like the in our case, different way of shopping. And of course, that's one of the things that people sort of see the real tangible benefit of AI agent is sort of acting on your behalf, supporting in your shopping experience. Of course, we can we can you know talk a bit more detail in terms of the technology.

Mosen’s Path From Hardware To AI Leadership

SPEAKER_01 4:46

Yeah, absolutely. But before we get into that, I wonder where did your personal AI journey begin? And and maybe just give us a brief kind of summary of where of that journey and uh taking you to where you are now as chief uh AI officer at uh at Kingfisher.

SPEAKER_00 5:04

Yeah, I mean that that's an interesting question. Uh I'm more than happy to talk about that. But before before addressing that, I would say, especially around chief AI officer as a title, I would say that title is more of a reflection of an organization maturity of the use of AI than a personal recognition. Of course, I'm I'm very humble to have that title and act in that level. But in terms of answering your question, I started my journey. I'm I'm coming from a very technical background, electronic engineering and microelectronics in my PhC and MSc. My PhD was in computer science when I was designing computer architecture, which was specifically designed to run high-performing computers, which is at that time it wasn't called AI. Machine learning was emerging, but the word AI wasn't that common. It was more machine learning baseball algorithm. So I was I was working on the designing the hardware that helped to run those type of algorithms at the scale. And then gradually, as you work around those, you know, I sort of I was I was very privileged to be part of a project called Human Brain Project at the University of Manchester, which Professor C. Ferrer was heading, which was the idea was how you they studied with the human brain process rather than just a conventional computer. So of course that's that gave me a lot of intuition in terms of you know uh how these things work and technically start dragging me more into the software side of it from hardware. And then since then I've worked in different industries, I've worked in special education using AI in computer or computer vision to understand kids with severe disabilities. I've worked for fashion industry, I've worked for you know uh pure digital businesses. Uh but in all of those different industries, I think uh the focus was all the same. So I I think it it was very important that AI wasn't the mission. So of course we talk about AI as a as a title, but there always was solving a specific problem that happened. I mean, that happened, AI technology happened to be a right tool to solve it. So yes, and I'm and I'm really enjoying it.

SPEAKER_01 7:15

Um retail, maybe not so much how you got into that or attracted you to that, but it seems to me that retail today it it feels just so dynamic. And perhaps it's it's uh one of the leading industries, if you like, in terms of using LLM and uh and AI, and or you know, we we say AI, but uh you know, I know you know far better than I do that there there are a lot of different uh flavors of of AI, and we'll we'll um we'll perhaps cover that in a second when we we talk about what you're doing at Kingfisher. But retail and AI and so forth, there must be a big attraction there in terms of what's going on.

SPEAKER_00 8:00

100%. I mean, personally, for me, I think physical stores add another level of complexity to the problem, which I quite enjoy. Like I like a challenge, and when it comes to retail and having physical stores, an extra challenge, they you know, of course, on opportunity and challenges that they bring. It was fascinating. That's how I actually get into. But you're absolutely right. The way I would like to describe retail is like an orchestra. So you have many different functions that they have to work in harmony for you to deliver, you know, your mission, which is serving the customer. And especially within our organization, Kingfisher with our brands, we have we have a very nice mix of you know different functions, different focus of businesses. You know, of course, you have supply chain, you have e-commerce, you have marketplace, and then within that organization, we have different focus like trade for scoop fix, we have big boxes like BQ. So all of those is all opportunity. It seems like a complex kind of organization, but that means opportunity. So, and that's why, for example, we have over 50 AI services in production, which is in the wide range of applications from customer facing to backend operation.

Building Teams Over Betting On Trends

SPEAKER_01 9:15

Now, something I saw that you you posted quite recently, uh which resonated with me, and I wanted to uh to discuss with you. You you posted something saying with a number of questions. What is the next big thing in AI? Which use cases should we invest in? Where does the biggest ROI come from, and so forth? And you said the truth is that no one really knows. These are the wrong questions. The most important investment is investment in people. And I know that the the group AI function at Kingfisher, you were telling me that uh a few years ago it was just five people, it's now 50. Talk to me a little bit about that and and and the the investment in people.

SPEAKER_00 9:56

Yes, and to be honest, the credit on this is mostly goes to our uh senior leadership, like Thierry Garney, our group CEO, Chris Blatchford, our CTO, and the mindset that you set up from a top line. And and the reality here is uh four years ago, when I joined King Fisher, we've been solving different problems than today. And those problems constantly change. The business, the challenges the business face change all the time. But what is still remain is we have same people almost, right? Or similar type of people solving different problems throughout the journey that we have. And that's that's what I think is the that's probably the biggest ROI is to invest and train and build the ecosystem around your people. Because if if you have a writer skill, then it doesn't matter what comes next. Nobody knew, or I mean, most general public didn't knew about the chat GPT release. So if you ask people a week before release of chat GPT, what is your priority and after release of chat GPT, so that just opened up new opportunities that people start thinking about solving different types of problems. So but in my team, the same people actually uh you know use just different technology because that's expertise and the skill set and academic knowledge and engineering knowledge. So, and and yes, and because of that, I think moving forward, you can navigate, right, whatever comes next. And and fundamentally all of those algorithms mathematically, you know, require the similar type of skill set to deploy and develop. So that's why I think the shift, yeah, we can keep predicting what's going to come next. Another reality here is, for example, if you look at Kingfisher and we have we technically selling the same type of product, but if you get if you go to different brands, if you go to BNK and Screwfix, you see their approach is different. Some of their strategic goal is different, their priority is different at a different time. So as a result, you know, it's not about the problem you're solving, it's about who actually have to be able to help with solving those problems.

SPEAKER_01 11:58

And I think you mentioned Screwfix there. I think uh it's fair to say that as a brand, it's probably best known now for the app and the fact that if you're an electrician or a plumber or whatever it happens to be, and you're on site and you need something, and this is not meant to be an advert for screwfix, but it's just a fact that you you're on site and you need a part, you'll get it within what is it, is it 20 minutes now that that's that's claimed something like that?

Speed And Convenience: Screwfix Sprint

SPEAKER_00 12:28

Yes, or our amazing yes, our amazing team at the screen fix. So they did an amazing job with the sprint, which is called, which is like technically 20 minutes delivery everywhere, right? You know, and and then you are and our team will keep working to improve it more, you know, and and it's all about that, especially for school fix, because the school fix the main cohort of customers are trade professionals, these are the people actually making money using DIY. And the time is the key, time is the money. So it's very important, and that's that's that's what I'm saying is right. So, your strategy, if you look at the actual strategy, then that the the time, convenience for our customer, this the technology to achieve that goal can be different, right? A few years ago, maybe it wasn't AI, today is AI, a couple of years maybe something else. But the focus is still the same. So the focus here is how you deploy this technology to make it cheaper, easier, quicker for your customer.

AI Replenishment For Trade Pros

SPEAKER_01 13:25

Yeah, yeah. Now, on that kind of theme, and I wanted now to spend a little bit of time learning about some of the initiatives that you're driving at Kingfish. And one that I saw was all about the AI-driven replenishment recommendations, which kind of goes to what you were just saying about time and and so forth, but convenience and being relevant to the customer and these sort of things. Tell us a little bit about that and and maybe some one or two of the other initiatives.

SPEAKER_00 13:55

Exactly. That and that's a good one because it is actually driven by our SQFix brand. So SQFix asked for that and pushed for that. That's because of the same reason it's the same reason you mentioned. We have a few different trade categories of electrician and and you know, and you know, different people actually working, decorators and many other different categories. And the idea here is those trade professionals regularly purchase some items, and some of those items are consumable. And of course, they are they're professionals, they know what they want, they know where to go and get it. It's not about not knowing what they want, but the idea here is how you can proactively make their life easier. And replenishment model was actually a series of series of AI algorithms, it's not really a big fat AI model, it's a series of AI algorithms that goes and identify replenishable product, identify the customers, different categories, and identify the purchase pattern and look at all those information, try to predict what they need in the next, let's say, 30 days. And they predict when they're gonna run out of a specific thing based on the purchase pattern, so therefore they can recommend it. And as soon as you have that knowledge, of course, you can recommend it, you can send them notification, you can send an email. And this is this is not about increasing revenue because they're gonna buy what they need, but it's about convenience, right? It's about making their life easier and quicker, making sure they have everything they need than they are on the side doing the project for our for their customer.

The Store As A Digital Hub

SPEAKER_01 15:26

Another one that I just wanted to ask you about, which again is obviously in the public domain, but and this is about the store of the future. And the and again, this is a theme that that came from out of NRF, and this is all about the store being a digital hub and connected and so forth. And again, I think it's something that that you've you've written about blending AI-driven personalization that you were just referring to. Where do you see the the store evolving if if you like? Because there's an awful lot that's talked about, oh, the store of the future, and so forth. Where do you see it uh it evolving?

SPEAKER_00 16:07

I mean, of course, there are I I'm first of all, I'm not the most expert when it comes to like SOR operation, but I was very lucky to be allowed to walk in a store and you know learn from our colleagues. So I think one of also as a customer, not just as a technologist, I think what what would happen, or what I think needs to happen, is try to close the gap between the store experience and online shopping experience, right? And of course, it's gonna be a journey. And uh one of the one of our first attempts was like when we introduced our Hello Casto agent, uh, which the idea was when you go to a store, normally you go describing your problem. You go and just say, I want to remove my old wallpaper, I want to fix this, rather than looking for a specific product. And within our stores, we have amazing colleagues, experienced, that try to help you and find that product and show you what you can do, which is more about instructions and just guidelines as well. On our online website, that wasn't possible at all two, two years ago, right? And the idea was on online, you had to go and search for a product. So the idea here is, especially if you are physical and we are lucky to have physical source, that we have that point of contact with the customer. So I think learning from a source and try to close the gap with the online and a store experience and making it a seamless experience for a customer, right? I think that's that's where we are heading toward. And of course, through that journey, a lot needs to happen, right? At the moment, I think we have a significant number of our uh orders or click and collect in B and Q, for example. I don't know the numbers on top of my head, but the numbers are many, many of our orders are click and collect, and that shows how we use a store as a hub to connect with our customer. And that opened up a lot of opportunities, so which you are purely online. So yeah, I think closing that gap by far probably is the most important.

SPEAKER_01 17:56

Yeah, yeah, yeah, yeah. Um I said at the start that we we usually uh we don't go on and on and on with the these podcasts, and I think we've already got some uh some uh gold out of this one. But if I can put you on the spot, well, I did this at the conference that we were at a few months ago, so I think you know what's coming. In in kind of like the last minute or so, give us your kind of closing thoughts and maybe where where we're heading, what the future, you know, and by the future I mean yeah, the next 12-24 months, what what we can what we can expect with um AI and retail.

SPEAKER_00 18:36

I mean, especially AI retail, of uh as you mentioned that early on in NRF, there was a new announcement in terms of some new piece of technology like UCP, which is Universal Commerce Protocol or ACP, which was exist before. Which is technically allowing AI agents to start shopping your behalf. So and we're going through that journey, and there's a lot of unknown and a lot of learning. So, and and I still think there's a lot of hype. I still think a lot of education needs to be done. So there is a development keep happening. We're gonna see more interesting technology coming up. But I think to help us to go through this hype and making right decisions, especially if you're a leadership level in the some of the organization, I think it's very important to learn and understand the consequence of using those technology and what are the for the customer, but also for the colleagues. So then we can deploy this technology in the right way to benefit both us and our colleagues, our people and also our customers.

Closing And Where To Learn More

SPEAKER_01 19:38

That seems a perfect place to to end on. So, Mosen, thank you so much. That's been absolutely fascinating. So that's all we've got time for from what's in the box for now. New episodes will be dropping every two weeks, so please stay tuned. And if you want to find out more about Boxtech, please follow the link in the description. Thanks to Mosen, and thank you to listening.

SPEAKER_00 20:00

Thank you, Anna.