Elegance At The Edge

Video link: https://youtu.be/oqY3eYD08G8

Editor’s note: this transcript has been lightly edited for clarity.

Olyvia:        
- Hi everyone, welcome to Avesha's July webinar titled, Elegance at the Edge: Disrupting Retail with Generative AI and Edge. Now all of us have been to a store. We've shopped online we've been to high-end retail, all other kinds of retail stores. Today, we'll discuss the tech revolution that goes on behind the scenes to empower these experiences. Today we are very fortunate to have Abi Sachdeva with us. He's the Head of Technology of Tory Burch. Abi has a wealth of experience in retail, and high-end retail, all forms of retail, and the topics that we'll be covering today. Abi, welcome, and thank you for joining us.

Abi:       
- You're very welcome, thank you for having me.

- And along with Abi, we have Raj Nair who's the Founder and CEO of Avesha. And Raj is a technology pioneer, and many of you may know this, he's the inventor of the first commercial load balancer that was acquired by Cisco years ago. So when you think of an industry like retail and the volume of transactions, and requests, and load balancing technology, it all started with Raj and Chang Woo. And I'm your host, I'm Olyvia. I'm a product, marketing, technology, entrepreneurship enthusiast, and its applications across all vertical markets. So let's get right in. Abi, Raj, welcome again. And we'll structure this conversation in three segments today. So we'll get into shopping experiences and how they're evolving. We'll talk about the technologies that empower those experiences, and we'll also cover some metrics. What does the industry look for when it measures success? So let's start with you, Abi. If you'd like to share about your journey in this space. And the second part to that is, how is luxury brands or high-end retail different from all other kinds of retail?

Abi:       
- Sure, about my background or myself, right? A little bit. I'm a kind of a leader who actually got into the retail by mistake, take it as that, in 2008 when the financial market crashed, right? At that time I lost my job, and I was sitting and talking to my friends, and then it's retail, Jen and Jenny from Rent the Runway came to me, and they said like, "We wanted to start a starter where we can do a rental of high-end fashion design and all that stuff." And that was my first get into. I was the first who built everything from the Rent the Runway and get into, and then since then, it's always been retail for me. So since 2008, now you can take it as that. It's been so many years. What drives me? It's definitely that fashion piece of it, I loved it, the retail side of it. And that's why majority of my experience in the retail is towards the fashion industry and all that stuff. In spite of a gifting industry, which I've entered into it for a few years with managing the flower. But other than that, it's majority in the fashion. What I like is like how the customer experience has to be built in the fashion industry and the retail industry. It's very different from what you actually can think of it from anything when you are just thinking from a technology side or a product side of it, because that's what you have to be very careful. And then the innovation in the retail, it's been crazy, right? How far we have come into, if you talk about ARs, VRs, and what we are talking about, we are going to talk about edge computing out here. We have done so many of those things, and how the retail industry has come from very monolithic systems to microservices, edge and all its flows into. And that's what drives me. And that's who I am as an innovator and as other people that I love to stay connected in this side of it. Now the second part of your question earlier, what you asked me was, how's the fashion retail industry, luxury fashion retail industry is different from the regular retail industry? The luxury industry, especially, is very steeped in the heritage and traditional side of it, right? If you notice, like many of the luxury brands still comes into it as they do not have an online presence, right? Or they have a sale which is coming very limited from an online side of it, because it's been such an exclusive experience, what they have built or what the fashion retail luxury has it, which is only used to be only possible in stores, right? When you go there, you get like a kind of white glove experience. They're showing you all that kind of a thing. And that's why it's very hard to replicate that experience in online, right? And that early, it was just the presence where the people were trying to have it. But now, literally, for a luxury value, if I'm not wrong, lately 30% of the sales, right, in the luxury fashion industry comes from the e-commerce, right? And that is huge in my mind, right? Because again, luxury, when you're spending like a $2,000 something, or a handbag or anything like that, you want to hold it, you want to feel it. When you are wearing that dress, which is exclusive, you want to try it before you buy it, right, so it has to be. But on the other side of it, what comes down to be every fashion, luxury or regular, is facing a 90% time the customer has touched the digital side of it, one or other part of their journey, right? Either they have experienced it into the store or when they're coming from the store and they're going back. So the presence somewhere has to be there. But regular retail industry, everything is e-commerce now, right? Its majority driver is e-commerce and all that stuff. And even from the luxury side, the omni pieces coming now are big, right? People has understood it, like the journey has to be together on that side.

Olyvia:      
- That's very interesting. So luxury with its aspirational nature is also moving to digital along with these in-store experiences. So Raj, how, how do you think, I mean, just starting out the conversation here, I mean innovation, we've seen that retail, or high-end retail, is, like we said, is at the leading edge of all these technologies. What are those initial technologies that such an industry has to adopt to influence these?

Raj:      
- Yeah, absolutely. I mean the NSRB had pointed out, you know, the move into micro services and having the flexibility to deploy them, you know, in a way that makes sense for your customers. If it is, for example, we touched on the new experiences and lot of that needs to be powered by AI, for example. And, you know, the better, the closer you are to the customer from a deployment standpoint, the better because you're going to get a better experience. So when you focus on that, then you need the infrastructure, you need the support from pieces that would allow you to do that. And the beauty of, for example, modern technology is, take Kubernetes, it actually allows you to do that. It is this universal currency. You can actually have it in different locations, and you can have your deployments closer to the customers. So this is some area where we have focused a lot on and we have developed certain products that make that possible, that Kubeslice. And we enable that ability for you to get a much better user experience, especially even in a store, like if you have a, for example, an AR or VR type of experience, this can be powered with modern technologies in this manner.

Olyvia:     
- Yeah, so we talked about two things. So experiences and AI, we'll come to AI, I want to focus on experiences first. So when you talk about feeling that $2,000 purse, or address, Abi, and you mentioned AR, VR, those are in store, or is it also digitally at home that customers want to be able, or you enabling customers to be able to experience some things?

Abi:     
- No, where the trend line is going, so if you look into it, majority of the luxury online shoppers now is generation Z, right, gen z guys, and they are the one. And if you look into it, all the latest technology trends, they're going to pick it up, right? So looking behind it and seeing it, like the feeling what they have to get it, right, when you cannot, you have to visually see yourself carrying that bag. You want to see that, right? And that's the majority of the reason people go to the store, I want to hold it and then I want to see how do I look with that bag, right? Or when I wear that dress, how do I look in that? And that's where the VR tech is coming into the life where it's ultimately where it's going to go towards is people going to have, people will be able to see themself holding, wearing that dress into that virtual world, right? In a very real, not with the avatar kinds of, avatars, it's still a turning point on that side of it, but in reality, where it's going towards, it's going to look very real, right? And that way it's going to be very valuable. Right now you can see yourself wearing a watch, right? As simple as that with an AR side of it, you can do that, but you cannot physically see yourself. And just in a mirror side of it, you can do it. But it's coming up, in my mind. Where you see where the trend line's going to go into it, it's started already with the smaller things, where it's physically going to go into it, it's going to be a pure AR, VR experience where the people will not need to go to the stores. And that's where I believe the luxury brands are adopting towards it. We have seen that window going into it because that way people can shop wherever they want to, and they don't have to be in that store to get that experience. Imagine that, just imagine for a second, right? You are in that VR world looking, very look like you wearing that dress, and then there's a customer service same as what you get in a store, which will be AI customer service, who will be talking to you there and then, right? Or actually a customer, so you're going to have a virtual store, right, with the AI tech, VR stuff, and all those things together, and that's ultimately where the game is going to change with this.

- And all accessible from the comfort of your home.

- All accessible from the comfort of your home. It's going to happen. I don't know the timeline for it, but this is where the trend line is moving towards it, especially for the luxury brands.

Olyvia:     
- Right, so, Raj, then this calls for an edge network that's really ubiquitous, right? It is across the whole country or maybe a global footprint.

Raj:     
- Absolutely, I mean that is the other innovation that is going to enable this is the ability to have an edge that is close by to where you are and give you that experience and deliver, you know, all of this, we had spoken about, you know, this VR, AR experience, in a timely manner with very smooth images and so on, so that you feel that. And it's very important because, you know, any little delays and things will detract from that experience. So you want to have that very good high-fidelity experience, and that's something that you can get with these edges that are coming up. And they're all very specialized, they are ubiquitous, as you said, right? Available in different locations. And then with technologies like Kubeslice you can actually federate them. So you can actually have this delivery to any of the, to the closest edge essentially, and with high availability. So that's the idea.

Olyvia:     
- What's the key thing in Kubeslice that enables this, you know, in a differentiated manner than what we have out there?

Raj:     
- Yeah, so the most important thing is that you should be able to run certain, so in a retail, you know, technology, it's not like you can run the entire stack in the edge because the footprint of the edge is small, and it's specialized. So what you do is you are able to run maybe the inferencing part, you know, the key part of it that is needed for that experience. You run that closer to the user and then the rest of the stack can be in a cloud, or in multiple clouds as the case may be. And yet you should be able to seamlessly connect, and you should be able to see let these transactions also happen, you know, smoothly. Because that's again, another important part of the user experience and that has to be very seamless.

Olyvia:     
- So all these modern applications become truly distributed.

Raj:     
- Yes.

Olyvia:     
- And so moving on to AI, so generative AI is, you know, everyone's talking about it, there are so many applications. What is it that you are seeing, Abi, in this space? How is generative AI going to, and you talked about the gen Z, or millennials, or the other new age, how is that going to be imbibed into shopping experiences?

Abi:     
- Customer experience is the very first one which I think about it, right? When you look into it, the customer experience is like literally, can be helped, and be enhanced with the use of generative AI, right? And that's there, right? Optimizing a store, like store operations also, we can actually be, we use. With the edge computing, we can optimize a lot of store operations and that's where the AI piece is going to come into it, right? If you think about it, how the location data is being used, right? That is also can be used on that side of it from AI pieces. These are the places where I see majority of coming in, but people have a very descriptive knowledge these days for generative AI. It's vast, and I think like that can take days to even to have a discussion about it to be how much that we talk, but I personally think majority is the customer experience, right? And the data it's going to be.

Olyvia:     
- Right, so these large language models will power the customer experience aspect of it. Now, innovation, we've taken a different spin to generative AI. Instead of predicting the next word or the next phrase, we predict infrastructure components. So, Raj, how are we using, or how is Avesha using generative AI in the infrastructure space?

Raj:     
- Yeah, I mean you're right, there is a similarity between what we are doing and what is done in things like Chat GPT, in the sense that, as you said correctly, we are looking at, so if you look at a system that you want to scale, and whether it is, so it's sort of the next level of load balancing, if you want to think of it that way. Well, what you're trying to do is you're trying to predict, you know, what is the next state instead of the next word. And so the importance of that is that then you can scale your resources in a proactive manner. So that is the fundamental concept. And so we use AI for that, and we have models that use also similar to Chat GPT, it was trained with reinforcement learning. We also use that with the human guidance. So with that, when we do it, we actually have a way of scaling that is proactive, that actually looks at user experience, because at the end of the day we look at application-level metrics. So we are looking at the feedback. So it may not be feedback from the user in, you know, verbally, but maybe it'll get there. But you know, maybe somebody gets angry and yells at this screen. Yes, you know, it could be, but on the other hand, we are looking at metrics of with the page load quickly, you know, the response time of that. And we are taking all of that in account to make sure that these workloads are running in the most optimum way and the most optimum location so that you can deliver these experiences seamlessly, and the infrastructure itself is responsive and meets your SLA or SLO as the case may be.

Abi:    
- And that's very interesting. And I'm sorry, just to add to it, Raj, what you just said, like how you can actually predict where the traffics and all those kinds of things and scaling up to that level. That's actually a very interesting point, how the generative AI could be used. And I was not thinking in that angle, but that's actually very interesting to look into it. One another place where I believe generative AI might be is the reporting, right? Is like how the businesses need so many different kinds of reports, right? If a very simple natural language person can ask, like, I need this data and the data is now in a common data lake, and once you actually have consolidated all the things, those reports can be created on the fly, right? With that we don't have to go into and look into how we do it right now, in a very SQL fashion to go write a SQL query, to write it, and then pull the report and all that. That SQL queries or any kind of those queries which need to be written can be written with the generative AI to pull and create the new report, which I just needed for something. Like, I want to need something for this product sales-wise looks like this, or why it's not been that something like that, right? So that could be very interesting to see.

Olyvia:    
- That's very interesting. So we are talking about AI helping with the insights. So, and then we are also talking about different experiences, so it's a lot of compute distributed across different systems, or geographically, and all coming together to provide either enhanced experiences or enhanced insights to technology leaders like you. I want to touch upon events like Black Friday, or Christmas sale, or you know, Boxing Day globally. So when the number of users grow up exponentially because you have an event, how does that impact what you do, and that kind of also relates to the generative AI solution, Smart Scaler that Avesha has?

Raj:    
- Yeah, so in our case with Smart Scaler, we have some really innovative technology where what we have done is we look at an application and we look at its service graph, and we are able to know how to scale this application in a manner that can handle a spike, for example, like an event, you know, you usually causes a spike and this is very, very common. So you have a promotion, for example, or you have a Black Friday or whatever be the event, you have to preposition resources. Well, commonly people don't know how much, how many resources they have to preposition, that's the challenge. But with our technology you would actually know that. And so you can actually predict, or you can preposition the right number of resources. And now this is the interesting part, like let's say that event happens, and to your point, Olyvia, it is even more than what you prepositioned, where we actually can help you scale all the way up as much as you need to get all of those transactions. And that's very important, especially, you know, because a lot of sales, and help me if this is correct, you know, a lot of these events like Black Friday, they're kind of important for retailers because a lot of sales comes from those events. And so it's important to handle and get all those transactions, and that's an important piece for retailers. And I'd like to hear what Abi says.

Olyvia:    
- Yeah, how do you, how do you handle that today, Abi, in terms of an event like Black Friday so that no shopper gets turned away, you know?

Abi:   
- We work very hard. But yes, the retail industry is all about Q4, quarter four, right? That's our like prime time. So preparing for that is like everything what we are trying to do to make sure we are doing all our testing first, our low test first, we simulate all those kinds of things to make sure we can handle that traffic and all. At this moment, we have not reached to the level where the edge computing can scale it to the level of what Raj was talking about it. But yes, we all are on cloud, so cloud cannot scale at this moment to a certain, not taking the full advantage of the edge we have where we should, but literally getting prepared, making sure we are scaled up, testing and all those kinds of things. And yeah, what we can't do is in a store, right? See the stores can't get prepared with all those kinds of things. Stores is like, we just need to make sure all our hardware tech stack, everything is running and updating. And that's where I believe the one of the things, what we was discussing, Raj, in the past was like having like some kind of an edge close to that store itself, where like if there is even an outage, something like that is happening, we still can take the transactions, inventory is there, we can work it out on those kind of a things. On the floor, there should never be a panic, right? Simple as that. And that is what we work mostly for is like there should not be a panic on the floor. Whatever we can do it is behind the scenes to keep the tech stack ready. That's the game.

Olyvia:   
- Is that what pop-up stores, yes, Raj?

Raj:   
- No, that was, going to ask, there was a question on that, go ahead.

Olyvia:   
- Yeah, there's a question here from the audience I think, “Are pop-up stores, extensions of traditional stores or extensions of the website, and what is a dark store?”

Abi:   
- I'd say pop-up stores are more of an extension. If I think from a nowadays world, pop-up store is closer to the online than to a previous retail, like a lifetime of like what you think into it, the extension of my physical store. The reason why I say the pop-up store is more, is like, because in a pop-up store, you're not going to have that much of an inventory sitting, right? It's just to show, like it's a small store, you're just going to see it in all those kinds of things. But you need to have your digital devices in your hand with all those kinds of things so that I can place an order. Even if I don't have that color available in my pop-up store right now, I know I have it online, I can just send it to you from there. So it's a kind of extension to your website now, right? Or extension to your digital presence. Then people used to think like, I'm opening of a stop pop-up store, which was a whole luxury, or any retail store, but now it's more of an online. I'm just showcasing my work, I'm showcasing my presence out there in that, wherever that pop-up store is opening, but actually order is going to ship it from there. And like I was reading, and the dark store is different. Dark store is a kind of a next step of a 3PL small warehouses. Let's take it as that, right. So every coming day, or not every, majority of the retailers who've been using 3PLs right now, but don't want to have that cost of doing it, they're going to have a small dark store, which will have its own inventory flowing into it. It's a kind of a small warehouse where there's a one person or two people who are just packing it up and sending it out. There's, the dark store concept came from the word where a customer is not allowed to go into the store, but it has the same point of sale system, the same kind of things going on, but it's more used for shipping. So think about it, if you can marry a pop-up store, dark store, you can actually have a very multiple locations of it where you're sending it, right? And you don't need that much of a customer service people to handle the customer, but your order taking process is all your digital. So it's flowing in from the nearest spot.

Olyvia:   
- Wow, and is luxury retail, luxury fashion, a global phenomenon now?

Abi:   
- Yeah, the luxury is likely. It's always been a global, right? You cannot control, you cannot be a brand just for a country.

Olyvia:   
- One country, yeah.

Abi:   
- It can't be, you have to have a brand presence like everywhere, and that's where it has to come into the window. You need to be known across the world.

Olyvia:   
- So then I'm thinking setting up all these pop-up stores and these omni child experiences and you talked about the supply chain with these dark stores, it's like, it's a huge setup to-

Abi:   
- It is, and you think about it, like you don't need to. If you go with the right strategy on the side of it, you have a pop-up store, and you have a dark store, and you have an online presence, and you market it yourself, right, any brand doesn’t need to be having a very expensive shops, like a retail space, to have a presence. You just need to be having the right customer, a place, simple as that just to showcase your brand. And then, the rest is all customer service. You deliver the product right, quality of a product is good, and you're going to have different story going on.

Olyvia:  
- Nice. Excellent. Raj, do you have anything to add in terms of technology? And I want to touch on security as well and we'll talk about that next.

Raj:  
- Yeah, I was just going to talk about that. But I think you beat me to it. The security is another very important aspect of what we offer, you know, from a place, because the slices not only they connect, but they also give you secure connectivity. And furthermore, if you wanted to isolate some traffic, two different groups of traffic or multiple groups, you can put them in different slices, and they will become completely isolated. So that's an important aspect again in today's world, you know, where you need to really security as number one, right, in all your decisions and planning. And we have and that is like a cornerstone of our product. But go ahead, Olyvia.

Olyvia:  
- Yeah, so in modernizing these applications and in moving to microservices, or Kubernetes infrastructure, Abi, would you say that the industry is, or the luxury fashion industry is 50% there, 80% there, or?

Abi:  
- Yeah, I think. I think like luxury fashion, it's like, as we are building an e-commerce presence, majority all the luxury brands are doing it, right, their first choice is going towards that side of it, the cloud computing, Kubernetes, and all those kinds of things. And that is where we all understand we have to do that, right? Going back to your data centers and all those kinds of things is not going to happen. Nobody going to go back to those ways of working. And security as a top of the mind for everybody, right? Like these things make sense itself, right? Because the last thing, what anybody wants is anything as a security challenge, right? So what Raj just mentioned, how it's coming up out of the box, getting connected, that's like, that's a game changer of its own.

Olyvia:  
- Right, so any, I know we are almost at the end of time here, Abi, any closing thoughts on how these experiences or technology would change, I don't know if I should say 10 years from now, it's hard to predict, but it'll be more of AI, more of these magic mirrors, and VR, and will be all very immersive.

Abi:  
- Yeah, I think like one, definitely in the near future we are going to see few things, right? Omni will be very, very strong, right? It is already there, kind of, right? Digital assistant, right? Think about, it's near, like it's coming, right? Digital assistant is coming, your personal digital assistant is coming who going to be like your personal assistant, either on the phone, or on computer, or wherever, who can shop for you, who can do anything for you, that's just around the corner, right? Virtual stores, virtual mirrors, AR, VR, we all know that, that we are walking onto the path, but nobody can predict future, right? Simple as that. There will be a lot more than this, but these are 100%, or near 100% happening.

Olyvia:  
- Nice. Excellent. So, and with these modern applications, so again, discovering these applications, talking to each other, ensuring the best experiences for the shoppers. So there need to be some parts of the applications on edge and networks, whether it's in the store, edge devices in the store, or closer to where customers are. So the scaling aspect, the networking, the discoverability. So I think those, those are challenges that need to be solved, which Raj was describing that products like Kubeslice and Smart Scaler address. Any closing thoughts, Raj?

Raj:  
- I think you're absolutely right. I think compute, from my standpoint, you know, we are we looking for a future where compute is ubiquitous, and you can just automatically call it up for delivering any of these experiences on demand and on the fly. And so that's where we're headed to and it's going to happen, you know, in probably even less than 10 years. But I'm very bullish about that.

Olyvia:  
- Yeah, thank you. Well, thank you both. Thank you, Abi, it felt very, went by very quickly, but you've educated us on a lot of the concepts and, you know, terminologies, and aspects that are going on in the industry, and technology and otherwise customer experience. And we look forward to talking to you again and going maybe in depth in some of these. And thank you for your time today.

Abi:  
- Oh, thank you for having me, right, and it was pleasure, this conversation was fun, thank you.

- Thank you.

- Thank you.

- Take care, guys.

- Take care.