In this interview from theCUBE + NYSE Wired: AI & Retail Trailblazers, Akash Gupta, co-founder and chief executive officer of GreyOrange, talks with theCUBE’s John Furrier about AI’s operational proof points. Gupta explains how GreyOrange’s AI orchestration spans supply chain and store operations, turning real-time signals into coordinated action across robots, people and sensors. The result, he argues, is an emerging “operating system” for retail that connects once-siloed nodes into a single, end-to-end decision engine.
Gupta walks through how AI reasoning can translate events and local context into tasks for associates and automated replenishment from warehouses. He also maps the next wave of physical AI, where robotics takes on the repetitive work and frees staff for higher-value customer interaction. Along the way, Gupta shares the scale behind GreyOrange’s models, from hundreds of millions of AMR miles to rapid growth across thousands of locations and a fast-expanding network of “agents” powering modern commerce.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI & Retail Trailblazers. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: AI & Retail Trailblazers.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Register For theCUBE + NYSE Wired: AI & Retail Trailblazers
Please fill out the information below. You will recieve an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for theCUBE + NYSE Wired: AI & Retail Trailblazers.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI & Retail Trailblazers. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: AI & Retail Trailblazers.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Sign in to gain access to theCUBE + NYSE Wired: AI & Retail Trailblazers
Please sign in with LinkedIn to continue to theCUBE + NYSE Wired: AI & Retail Trailblazers. Signing in with LinkedIn ensures a professional environment.
play_circle_outlineImportance of integrating physical and digital experiences in retail environments.
replyShare Clip
play_circle_outlineTransforming Retail: How Consumer Behavior Analysis Shapes Personalized Shopping Experiences and Dynamic Store Layouts of the Future
replyShare Clip
play_circle_outlinePhysical AI and digital AI agents improve efficiency and customer engagement.
In this interview from theCUBE + NYSE Wired: AI & Retail Trailblazers, Akash Gupta, co-founder and chief executive officer of GreyOrange, talks with theCUBE’s John Furrier about AI’s operational proof points. Gupta explains how GreyOrange’s AI orchestration spans supply chain and store operations, turning real-time signals into coordinated action across robots, people and sensors. The result, he argues, is an emerging “operating system” for retail that connects once-siloed nodes into a single, end-to-end decision engine.
Gupta walks through how AI re...Read more
exploreKeep Exploring
What is the context and purpose of the discussion in theCUBE's NYSE studio related to the AI Trailblazer series in retail?add
What is GreyOrange's perspective on their role in AI orchestration?add
What are the implications of physical AI on retail operations and customer experience?add
What are the challenges and potential solutions for implementing changes in store layout in a retail environment?add
What role do physical and digital AI agents play in enhancing the retail environment and consumer experience?add
>> Welcome back. I'm John Furrier, host of theCUBE. We are here in theCUBE's NYSE studio. Of course, we have our Palo Alto studio connecting Silicon Valley and Wall Street, tech and capital. Technology is the market. And this week, retail week is happening. This is the AI Trailblazer series in retail where in retail, which is an edge environment, you're starting to see the physical and the digital worlds come together. This is where the experience's changed, the business operations change, everything's changing, again, just one of many verticals of retail trail. We have one here in the studio, theCUBE alumni, Akash Gupta, who is the co-founder and CEO of GreyOrange. Akash, great to see you. You're back. It just seems like yesterday. It was almost a year ago.
Akash Gupta
>> Absolutely.
John Furrier
>> Welcome back.
Akash Gupta
>> Thank you. Thanks for having me again. And things are moving really fast.
John Furrier
>> We got a little upgrade here. The trading floor is going crazy. NRF sets the agenda every year right after CES. It's almost like a one-two punch. You got CES, this year was great with Jensen up there, physical AI. But physical AI is the next wave of retail, obviously brick and mortars, but a lot of the internal plumbing is getting done in retail. The data plumbing, the models are starting to get adopted. You're starting to see agentic with real proof points. I don't think there's a strategy risk in retail. Infuse AI in all the business. That's a strategy. People get that. It's the execution risk we're seeing. Proof is in the pudding. We want to see proof points. This is the market. You guys are in the middle of it. You got an edge solution. What's changed for you this year?
Akash Gupta
>> Yeah, I think, firstly, a lot of advancements that has happened is starting to get together. It's starting to get real. A combination of compute on the cloud and compute on the edge is starting to make things very, very real in the stores. In the stores for the consumer, in the stores for store associates, in the store for visual merchandisers, and likes of that. So I think this year and last 12 months we've started seeing things, on concept level, getting real and, from that point of time, getting to real dollars and getting to real customer experience.
John Furrier
>> You guys have a lot of traction. We've been following you guys for a while. GreyOrange. People that aren't familiar with what you guys have. You got some good go-to-market momentum. Product's great. Give an update. Set the table. What do you guys do? What's the current state of the art?
Akash Gupta
>> Yeah, so from GreyOrange's perspective, we are basically a AI orchestration play. Unique thing about us is we basically orchestrate from supply chain to retail stores. So we have our software orchestrating robots, people, technology and warehouses, and then software orchestrating people, technology sensors in retail stores. And that's the power of autonomy era that we are going in where you can really consume large amount of data sets, connect multiple nodes and make sure that you can optimize things end to end rather than looking things at silo. And we are right at that point of getting to a point where commerce can be one, whether it is physical retail, online retail, you can look at it all at the same time.
John Furrier
>> One of the things I've observed, and first of all, I love the fact that you guys have this systems' mindset. In retail, one of the things that's very clear this year, and I think there's no debate, but maybe last year there was is that retail grew in kind of an IT sprawl fashion. Email system runs over here, email marketing there, you got automation there, you got the database. Now the word orchestration comes into play. You speak of orchestration at a different scale. Could you scope your level of orchestration, because I won't say harder, it's a little bit more complex? Compare and contrast what the word orchestration means on your side and how retail can inject that into their business.
Akash Gupta
>> Sure. Yeah. Orchestration basically means two things. The combination of two things. One is consuming all the data set, consuming all the data set in real time and understanding what decisions to be made, right? And a lot of people do this, but once decisions are getting made with that intelligence, how do you get to execute them, right? So orchestration is planning and execution together in real time. That's what orchestration means, which means that, I'll give you an example, very recently we have been working with a customer. So let's say you looked at a particular zip code and said, "Okay, there's a concert happening on that zip code," the intelligence layer basically orchestration layer looks at that and says, "You know what? For this, my sale window on retail store should have this on them."
You basically execute that intelligence by giving a task to the store associate to say, "You know what? Concert is happening. I want these two apparel items on my sale window." You don't stop there. Now when you know that's going to happen, you also trigger your supply chain and saying, "You know what? I need a hundred quantity of this from warehouses to store because this demand is going to go up." So basically look at the complete world scenario, take a decision, get it executed, end to end from store to supply chain to manufacturing and likes of that. And that's what's happening with GreyOrange in retail this year.
John Furrier
>> And I would just point out just for the people trying to crock this concept, the reasoning plays a big role in this.
Akash Gupta
>> Yeah, yeah, absolutely.
John Furrier
>> So AI actually enables this. I mean, connecting systems has always been kind of hard integrating, but so tech challenge, so reasoning comes in, you have to go, "Oh, what kind of genre is it?" Heavy, metal, EDM or whatever, country? I mean they all have different buying patterns-
Akash Gupta
>> Absolutely...
John Furrier
>> kick off. What products do they buy? Is it raining?
Akash Gupta
>> Yeah, yeah.
John Furrier
>> Might need some umbrellas.
Akash Gupta
>> Yeah, absolutely. And that's been the power of AI that you have that chain of thought reasoning that can come in, which means that you're not setting rules anymore. You're basically saying this concert, this demographics, this location, this inventory and it's all coming out together and saying, "Okay, you know what? This is what's going to happen." And you can execute that, measure that and then take a better decision next day. And we are seeing stores and e-commerce websites almost having a life of its own. They're evolving of its own from all the data, all the reasoning, all the understanding.
John Furrier
>> It's intelligence, it's got its own neural network, it's nervous system. I wrote a post on SiliconANGLE going into an RRF with titles called Retail 2026, when AI becomes the operating system. And you're kind of pointing at this because I want you to take me through the use case because it sounds easy. "Oh yeah, I got to see all the data," but you're talking about different types of data. You got warehouse, you got distribution, you got supply chain, you have different elements. Talk about the complexity. How hard is it, how much end-end do you need? What's the minimum? And then where does that flex? Can you take us through the deployment?
Akash Gupta
>> Yeah, no, absolutely. I think firstly, when you try to train these kind of world models like store world models or warehouse world models, you need a lot of data because synthetic data can only take you to a certain point. You need a lot of data. I'll give you an example. When we are training our store and warehouse world models today we have data of 250 million miles of AMR movements. We are basically consuming a hundred million of units of EPCs, RFID tags every hour. So you need that level of real data.
John Furrier
>> That's huge scale.
Akash Gupta
>> So you need that kind of scale and then you need understanding of how to create embeddings for each of these data types. How do you contextualize all of that and how do you finally take all of that data, go through this chain of thought reasoning and then create task and get executed? And when you're creating task and executing, you also need to understand what is the skill set of the agent that is executing it, right? So let's say if there's a task that needs to go up, you have to choose a forklift versus for a simple task, you can choose a person. So basically first basically take all of the data of different types, contextualize it, go through the chain of thought of reasoning, and then create task, understanding what kind of capability you have in which of the node and then get it executed.
John Furrier
>> That's why I like Jensen two years ago at GTC is like static versus dynamic or generative. You don't know what's going to come next. You got to be ready for that through the intelligence layer comes in. Talk about the platform because based upon some of your momentum with your partners, you're seeing success. Where's the layer of handoff? You guys view yourself as an enabling platform? You mentioned Zebra Technologies, we've covered them in depth, they have a lot of customers. So are you guys shipping a platform to them? Where's the handoff? Because domain expertise becomes big, obviously having scenario planning or execution and with reasoning it requires some sort of beauties in the eye to beholder. The data is the data. Where do you kind of stop and hand off? Where's that? Take us through that.
Akash Gupta
>> Yeah, so basically we take over once a user, a consumer orders something on a website or enters a store or likes of that, that's where we kind of take over the experience of the consumer and then we execute it to the point we hand over the good to the consumer or to a third-party logistic like a FedEx or UPX for shipping and things like that. So basically helping a customer generate the right demand through giving all the understanding and intelligence and then making it to the point where either that good is shipped or it's handed to the-
John Furrier
>> So you would consider yourself an operating system in my example of my post?
Akash Gupta
>> Absolutely. Absolutely. GreyOrange is basically that orchestration layer operating system for retail, connecting all possible nodes together.
John Furrier
>> Okay. I want you to explain, this is an expert question to you because I know retail, operating system in my title is really more of a lean towards NVIDIA because they love computer science, but for a non-computer science person in retail, explain what an operating system is from your standpoint because I think this is going to be more and more of the discussion.
Akash Gupta
>> So I think let's try-
John Furrier
>> From their standpoint.
Akash Gupta
>> Yeah, let's try break down operating system. Basically how retail has to operate is basically there is a certain amount of demand that is coming in. Then you have to take all the decisions and you have a certain amount of resources. Think about operating system in a computer. Basically what you're trying to do is there's an application which is trying to execute something. You have certain amount of memory, you have certain amount of disk and likes of that, and you have interfaces, right? Interfaces to take inputs which are like keyboards, mouses,
John Furrier
>> You got subsystems you got-
Akash Gupta
>> And likes of that. So you have basically multiple systems which have their own constraint and you're trying to achieve something out of that thing. And same as with the retail. So basically operating system takes care of all the decisions need to be taken and then gets it executed by real devices, whether it is showing you-
John Furrier
>> It runs.
Akash Gupta
>> Yeah. And it basically kind of does it at very, very high frequency that we all get doesn't even realize all of that needs to happen in retail. Millions of decisions are to be taken in retail every single minute. What to manufacture next, what to order next. Okay, now I've ordered, it's manufactured, how should I distribute it once it is distributed coming to the store, where should I keep it? How much should I price it? Should I keep it here at the front or the back? Which two products should be kept together? All of those decisions need to be taken and then executed.
John Furrier
>> We've all had experiences in our lives where you buy software. It's like it doesn't run on that operating system. Okay, so at the end of the day, the application in the OS example in retail, it's what they want, what their business is. I want to ship products to a customer, I want to provide an application or a service. This becomes the thinking. So I have to ask you this, in your opinion, given what NRF's evolved to, do you think that the retail industry is thinking more systematically in their holistic view or is it still kind of like that IT sprawl and then trying to put an abstraction together? Are tech leaders and business people thinking about the operating system and then the applications that are minimum to needed to run? Are they thinking that way in your mind or are they just in the first inning?
Akash Gupta
>> I think it's still on the first inning. I think it's still very early days and retail has gone through, this is how I think about this, right? Retail was going on, everybody was trying to innovate on retail and then suddenly e-commerce came in and all the focus went to e-commerce and the physical retail part of it got ignored and ignored for a decade or so.
John Furrier
>> Brick and mortar's dying. It's exploding.
Akash Gupta
>> Yeah. Exactly. And that's still like 70% plus of the overall retail commerce, right? So I think we are starting to get to the point where people are again back on, "Okay, you know what? Okay, e-commerce is a thing, retail is here for to stay and how do we bring all of that together?" So I think we're still in innings of that. There's some way to go.
John Furrier
>> All right, so I have to ask you about the physical AI piece because you bring up a great point. I agree with you by the way. I think there's always like, "Oh yeah, it's not the trend. The zeitgeist says digital everything go e-commerce, retail's dying." Obviously the connectivities are an issue in some places, but I think that's now gone. I think people realize you can have a physical store or outlet or location and have a first party digital converge that's converged.
Akash Gupta
>> Absolutely.
John Furrier
>> And you see crypto, all these other mechanisms clearly points to that sign. So as physical AI comes next, agentic will dominate in the short term, they'll recast the plumbing, get systems like you guys in there, get gray matter, GreyOrange in there, all that stuff will be deployed and then they start thinking, "Okay, how do I get the power to the retail?" Where does physical AI come in? Because agents will get the job done, you'll see some benefit, but the real investment's going to be how do I physically turn and experience face-to-face or physically marry that into digital? The edge is in place, technology at the edge, it's technology while you're immobile. So I'm in New York, I can be in Central Park on my phone, walk over to Nordstrom's if I want to or not.
Akash Gupta
>> Yeah. See the good thing about retail is retail is basically a game of bits and atoms. It's not about buying a plane ticket that is all only about atoms bits and you kind of go and do it, right? Retail is a game of bits and atoms, whatever you do, finally you have to deal with a physical good moving from point A to B to C to a customer and then to return and then back to the warehouse. So that's really, really relevant for physical AI. And what physical AI is enabling is different devices, robotic devices and sensor devices and likes of that to be very, very intelligent at the edge. And that's basically going to change how things operate. I think we are definitely going to see a lot of store operations, which are I would say not very value-added for the customer experience going to be kind of robotics, all the work that happens in night and the third shift, I think that's going to all move to robotic systems while the human interaction becomes the center of it, where you really have a well-educated AI powered human interacting with you when you're going to the store and likes of that. So I think physical AI is going to make a lot of room for human to human interaction, which we value today when we go into the store and take away all the boring part of operations and get it automated.
John Furrier
>> I think you're on something really big there because I think what the physical AI will bring is relevance to the physical thing, product, person, real world or robot. The edge has to be converged in my opinion. I'm sure you agree with that. But once that happens, once an AI factory is in the store, I mean a DGX box from NVIDIA is this big. I mean, it's tiny. Plug that in, connect the access points, collapse the spectrum, an unlicensed spectrum and spectrum together. You have a unlimited interface, that's going to happen.
Akash Gupta
>> Absolutely.
John Furrier
>> What happens next after that?
Akash Gupta
>> Yeah, I think after that basically you get the power of giving your customer entering in the store a completely personalized experience. Not just that, you get the power of setting up the store for that particular zip code in the way that demographics. Like today when you set up a store basically on-
John Furrier
>> It's cut and paste.
Akash Gupta
>> Yeah, it's cut and paste. Think about this a Fifth Avenue store and a SoHo store are set up the same way. The demographics are completely different. Once SoHo is being listed by tourists and Fifth Avenue with completely different people, you still set up the store same.
John Furrier
>> And the reason why they do that is cost.
Akash Gupta
>> Yeah.
John Furrier
>> It's simply mechanics.
Akash Gupta
>> Exactly. You just can't have so many people to think about this. You don't have that level of data in likes of that. Once you get all of that data and the intelligence that we are talking about and then power of execution. Let's say this digitally AI agent tells you, you need to rearrange the store this way. Where will you get them man-hours to do that? Retail is a low margin business. You can't really spend a number of hours. Now with physical AI and robotics and likes of that, you can change your layout every single day without worrying about it. So your next day, your store is set up for what happened yesterday and that's all going to happen.
John Furrier
>> Yeah, I love what you're talking about. And also I will point out that you guys, you in particular and your company have been on both our robotics AI series and now retail, this crossover there rightfully so, because the far edge, whatever you want to call it, if the edge is the access to the store environment, the edge, edge is the human or robot.
Akash Gupta
>> Absolutely.
John Furrier
>> And that software will be tied in a network configuration that's just more value.
Akash Gupta
>> Absolutely. See, I strongly believe retail is all going to be all about digital AI agents and physical AI agents because retail, as I said, is a combination of bits and atoms. So retail is going to be governed by somebody who can master physical AI agents and digital AI agents all coming together to give that right consumer experience.
John Furrier
>> Explain what that mastery looks like. In your opinion, I know it's early to flesh this out, but if you had to give a quick speech or master class or riff on that, what would it be?
Akash Gupta
>> Yeah, I think when you talk about digital AI agents, it's all about being able to use different kind of neural networks and models in the context of a situation. Basically, I always try and go down to basics. So let's say if you ask me a question, right? Now, I have so many options, I can think about it, I can Google about it, I can ask Claude, I can ask GPT and likes of that. I need to decide which is the best way of getting that information. That's exactly what agentic means. Basically our agent has access to lots of tools. It basically understands what's needed, acts as a combination of these tools and gives you what needs to be. That's digital agents. Once that is done, when you're talking about retail, basically it'll always involve either talking to a customer, moving a particular item from point A to point B or getting it shipped from the warehouse. That's where physical AI agent comes in. But that also needs understanding of what happens when you touch a product, the sensor data, how do you coordinate all of that. So physical AI agent is all about understanding the behavior of that hardware and embodying intelligence inside that to execute something.
John Furrier
>> And also you mentioned the store layout as an example, a great use case. Let's just kind go in the future a little bit. I mean it's attainable, but his would be the scenario. The robots, they don't sleep. You could have a store looking like something when it closes and be completely different the next morning.
Akash Gupta
>> Absolutely.
John Furrier
>> If it's programmed properly and you had the right workflows.
Akash Gupta
>> Absolutely. And I think it's going to happen faster than what we are all imagining because what I've seen in the last 12 months, if you ask me this question in March, I would've said, I think it's like three, four years still to go. But I think it's going to happen in a couple of years. I think we are going to see such use cases happening in a couple of years. The rate at which the embodied AI and things like that are kind of going forward. I think all of this is going to happen very fast, where next day the store is set up from all the understanding that it got last day.
John Furrier
>> Yeah. Okay. What's next for you guys? Give some stats on momentum since we last talked in March. NRF obviously popping, that show's great. I mean, 45,000 people, 40 to 45,000. Massive numbers. I mean, retail is a tech show. Technology is the market now.
Akash Gupta
>> Yeah. Yeah. Yeah. Absolutely
John Furrier
>> I mean this is happening.
Akash Gupta
>> Yeah. I think if you look at last 18 months, GreyOrange has added 3000 physical kind of locations that we are powering through AI orchestration. We have added a hundred thousand agents and for us agents says anybody who we are orchestrating, it can be a robot, it can be a sensor, it can be a person and likes of that. So literally in the last 18 months, we have added a hundred thousand agents to our platform that we are connecting with, talking with and orchestrating. And that's faster than what happened in last five years.
John Furrier
>> For the business people out there. Who's your customer? Who do you sell to? What's the value proposition?
Akash Gupta
>> Yeah, so I think for our customers, it's basically any retailer out there, right? From H&M's of the world to Fabletics of the world, to Ikea's of the world. So basically any retailer who deals with e-commerce, retail store warehouses is our customer.
John Furrier
>> And the buyer is technical or business or both?
Akash Gupta
>> Both. We sell to chief supply chain officer, we sell to CIO, we sell to CEO. Because if you're trying to build that operating system, all of these people need to come together and say, "How do we get that basic fabric of your retail by getting right tools. And so think about us as fabric of physical retail that needs to connect with
John Furrier
>> It's a platform.
Akash Gupta
>> Yeah.
John Furrier
>> Total platform and you plug in partner with folks out in the industry.
Akash Gupta
>> And on the other side-
John Furrier
>> Like Zebra of the world.
Akash Gupta
>> Exactly. On the other side, we partner with robotic companies who are building robots, which we orchestrate. We partner with people who are doing handles. We partner with people who are doing cameras because for us it's all about consuming any kind of sensor data and then executing through any kind of agent.
John Furrier
>> So it'd be safe to say that you guys are at the nexus of physical AI and digital AI.
Akash Gupta
>> Absolutely. Absolutely.
John Furrier
>> All right. What's next issue? What's the focus?
Akash Gupta
>> The focus for us is to start getting towards more and more autonomy. I think there's this debate of getting from automation to autonomy, right? Automation is when things can be done without human intervention. But the decision making is done by human. Autonomy is when the decision making is done through AI and gets executed through autonomous agents. So I think we are basically pushing the world towards autonomy as much as possible.
John Furrier
>> Okay. It's great to have you on again. Good to see you. Congratulations. I love what you guys do. Again, you guys cross over robotics leaders as well as retail. Again, the technology is the market. That's our vision of theCUBE. Thanks for coming on our retail AI Trailblazers. Keep blazing.
Akash Gupta
>> Absolutely. Thanks for having me again.
John Furrier
>> All right. I'm John Furrier. Again, AI is changing the world. And then retail specifically, that's a physical and digital converged environment. User experience, personalization, business operations as well. Don't forget that, that's what's happening. Again, this is having an all-verbal retail in particular will impact our experiences and of course how we interact with each other and the products that we love every day. theCUBE's doing it's part, bringing you the data here. Thanks for watching.
John Furrier