theCUBE + NYSE Wired: Physical AI & Robotics Leaders | Matt Rogers, Mill

Clips
More from theCUBE + NYSE Wired: Physical AI & Robotics Leaders

Matt Rogers

Co-Founder & CEO

Mill

play_circle_outline Mill’s Residential AI Dehydrator Comes to Amazon and Whole Foods, Turning Food Waste into Procurement and Menu Insights

play_circle_outline Edge-AI Trash: Computer Vision and Sensors Measure Food Type, Weight, and Water Content

play_circle_outline Uses agentic AI to connect siloed systems and automate cross-system inferences

play_circle_outline Market focus: grocery, restaurants, stadiums, universities, hospitals; canonical customers targeted

Info

Matt Rogers, Mill

Matt Rogers

Co-Founder & CEO Mill

search

John Furrier

>> Welcome back. I'm John Furrier with theCUBE. We are here in our NYSE studio. Of course, we have our Palo Alto studio connecting Silicon Valley and Wall Street technology is the market. All the stocks, all of our lives are infused with technology, the smartphone, cloud generation. Now the AI native generation is upon us. The AI infrastructure is expanding very, very fast. And again, we'll continue to tell that story of AI factories, AI factor to the edge. Of course the real world and digital world is coming together. Matt Rogers here, CEO of Mill, a company tackling a unique problem, but it's kind of a climate change opportunity, but it's also a very clever device. Also, the co-founder of Nest. Great to see you again. Thanks for coming back in.

Matt Rogers

>> Great to be back. Thanks for having me.

John Furrier

>> So I just love your pedigree, if I can say that word, Nest, Apple, you're tinkering, you're building, you're engineering your new venture. Give a quick update because you were in early. I know there's a lot of updates I want to get to quickly, but give a quick overview of what you guys are doing.

Matt Rogers

>> Yes. I'm building a company called Mill and we're building a waste technology company where we're combining AI and hardware to prevent waste and build a new waste system. We started with a product at home, a residential food recycler, which helps people manage food waste at home. We recently announced just a few months ago a new commercial scale version where we're deploying this first to Amazon and all their grocery stores and Whole Foods locations where we're also building an AI, a camera technology that's going to identify what food's being thrown away. It's a pretty holistic product that not just is helping manage waste but prevent waste altogether.

John Furrier

>> So check out that video on YouTube last time you were in. It was a home product, but it was very interesting because you're using the technology to do all that in this so there's not a lot of waste. Now the commercial business, like businesses who throw out a lot of stuff, restaurants, food, shopping malls, all that's getting wasted. You have technology. Explain the secret sauce to what you guys do.

Matt Rogers

>> Yes. So about half the food we waste comes from homes, the other half from businesses. The core of our technology is dehydration. So most of food is actually just water. So you think about those garbage trucks that are driving in our neighborhoods, 80% of food is water. So this is a lot of water being trucked around. So the core of our product is we take the water out of food and when you dehydrate it, it gets small, almost looks like coffee grounds. Doesn't smell, doesn't go bad. Also, you don't have take the trash out as often, but if you think that as a business, not just are you picking up the trash less frequently, but if you add AI, could you identify what food is being thrown away, when it's being thrown away and why? And then could you feed that back into your procurement systems, your menu planning to prevent food waste altogether and really save money?

John Furrier

>> I really liked the practical approach you took on building this company. One home, half the problems there. So tackle that first, get the product ready. It does what it does technically, solves that problem. Now you say, okay, let's take a bigger problem, commercial. Give us the updates. You got AI in there, you mentioned before we came on camera. Explain the advances of now a connected device could be better than a non-connected device.

Matt Rogers

>> It's pretty amazing what could be done today. So I started Nest about 15 years ago and for those, remember we had these Nest cam products, video doorbells, things like package detection where the UPS driver would drop off a brown package and you get a notification at home. Now we built that, we hand-tuned that model. It took like a year to build that. Dozens of engineers. Now with latest LLM technology, we could do this so much faster. What used to take years now takes weeks and the capacity to train these models, it's way faster. It's remarkable what could be done today.

John Furrier

>> So last time you were on last year, now this year. So what happened? You go back to the ranch, you go back to the team, AI is now hitting the table. We're starting to see those advances, acceleration. Jensen Huang calls it accelerated computing. You got super computing capability, you can buy it anywhere. GPU clouds, Amazon on a device. What did you guys do? What were the meetings like? We have to get AI in this thing. And well, take us through that.

Matt Rogers

>> There exists products and services out there today that try to help restaurants and grocery stores prevent waste. But usually they're additional work. There are things you add into your workflow and we were thinking is what if we could build it into what you do already? And today people take their foods, scraps, they put it in the trash can. Now what if we could intercept it just then, just when they're in the middle of throwing things away, not adding more time to their work? And this is exactly what AI is great at. And in this case, as you said, computing has gotten so good. We actually built the data center into the trash can.

John Furrier

>> Now that's an edge device. All right, so what's going on? So the computer vision, is there a computer vision? How does it detect?

Matt Rogers

>> That's right. We use a network of cameras plus sensors and we kind of combine all those together to get a really accurate picture of what's being thrown away. Because the vision technology is really good identifying what is being thrown away, but you need other sensors to determine how much they weigh and how much water is in them and that kind of stuff.

John Furrier

>> All right. Take us through the expansion into commercials. I find this fascinating because, and I'm curious how the AI is reporting the data. So you've got Amazon and Whole Foods, they have an operation, they have inventory. People can throw stuff away if they're done eating at Whole Foods, but they have the buffet there, they got tons of food, they got produce and they got tons of food that's perishable.

Matt Rogers

>> That's right.

John Furrier

>> Is that the waste? Take us through the problem and what you guys do to solve it.

Matt Rogers

>> So the core problem is food waste is several percentage points of revenue for these companies. And for those who know the grocery business, grocery business is not a high margin business. So a couple points of revenue is a big deal. So if we can go and identify what are the sources of those waste, when are they happening and agentically, what can we do to prevent that waste? Tying in automatically to their procurement systems or their menu planning systems. What times a day are these things happening and could we automatically make these inferences to prevent the waste from being happening?

John Furrier

>> So you take the edge device, which is the trash system with the data center and all of that. My mind's going crazy there. It's a rack of servers. And then you're connecting to essentially databases which are traditionally siloed workflows.

Matt Rogers

>> That's right.

John Furrier

>> You're tying that together. Is it agentic? Are you building in coordination, delegation? Take us through some of the things going on.

Matt Rogers

>> Exactly. This is an ideal use case for AI. This is actually exactly what agents are for. The kind of things of connecting these dots that are really hard to do manually and in the spirit of what's happening in the software world today, this is one where if we did it the old way, if we built custom SaaS solutions for every vertical, whether it's a custom software solution for quick service restaurants versus grocery stores, we would need thousands of engineers to do that. And this is kind of the ideal use case for agentic AI.

John Furrier

>> So I have to ask you Matt on, while you're here, since you're an expert. We have a mixture of expert series. So we'll kind of throw that in too. Yeah. The whole SaaS getting killed by AI conversation. I interviewed OpenTable last week, they were here and they been around for 20 plus years. They got tons of venues. Took a lot of work to get that supply chain down. They got the demand side with the user base. They're a SaaS app, they're like the textbook cloud app. I mean you get a restaurant order. I don't see them going away. So I mean, so where do you see the line between a competitive moat, either technically, system of record or data versus say the interface changes where things are being embedded in like you're doing with the waste systems or prompts or now the interface? What's your take on how you see, I'm sure you get this all the time, is SaaS getting killed by AI? What's your-

Matt Rogers

>> John, you know I've been doing this a long time.

John Furrier

>> Yeah.

Matt Rogers

>> We've seen the arcs of technology and think back to 1999 when Cisco is the most valuable company in the world. Right. The rise and fall of technology companies is one that's tale as old as last several decades. I think about companies like Nokia and RIM and what happened with the iPhone. The question with these SaaS companies is now the barrier to entry has gone down and custom solutions can be built much faster. How are they evolving their businesses? And one of the things that Steve Jobs taught us at Apple is you got to think about not just your competitive moats, but can you go after your own businesses?

John Furrier

>> You mean compete on value.

Matt Rogers

>> Compete on value. So-

John Furrier

>> And cannibalize your own business with better product.

Matt Rogers

>> You got it. So we were worried about this. This is kind of the genesis of the iPhone. We were worried. Hey, smartphones are coming. They're going to eat our iPod business, which was most of Apple's revenue at the time. So part of the genesis of the iPhone is kind of defending the fort on the iPod business. Now I think about it, if I was a software CEO right now, I'd be thinking about that. What are my competitive moats? Where's the puck going? And then start building solutions to where the puck is going.

John Furrier

>> And so that is classic business strategy product. Product market fit. Question isn't are we going to be around, the question is product market. That's what you're saying.

Matt Rogers

>> That's right. In this case the market is shifting and product is shifting at the same time. The competitive moat that a lot of SaaS companies had of it would take tens of thousands of engineers to build this product. That barrier has gone down.

John Furrier

>> Yeah.

Matt Rogers

>> So they have to think about their moats differently.

John Furrier

>> And that is the interface. All right, so back to where you're going with this because you're now again, one of my favorite topics that I'm working on is hyper-converged edge. You are a working example of a device perfectly positioned for being an edge device. You got compute in the device itself or the system. It's a system, it's not like a device, it's not a tool.

Matt Rogers

>> That's right.

John Furrier

>> You're building essentially an OS within the system.

Matt Rogers

>> Yes.

John Furrier

>> It's going to connect in. Eventually the model's going to want to come down to your device and make you smarter. How do you see that going on your vision? Because as the world gets smarter, intelligent wise, you're just another node in the network, in this case multiple nodes. Amazon has all over . There's their network. Then there's the external network coming together. How do you see that architecture emerging?

Matt Rogers

>> Yeah, I think actually even back to my time at Nest, my co-founder Tony Fadell and I envisioned this even back then. And at the time technology was just not good enough. We actually wanted to be able to build all the computer vision into the Nest cam itself and we were just not there yet. But we're at the point now where I think it can be done. And I think from a privacy standpoint, but also just like compute resources and energy perspective, it's much better to do these things locally. The more local the better and being able to just download a new model or take the inferences and learnings from all of the nodes of the network and build new updates to the model and push that back down to the edge. I think this is the way of the future. I think this is where things are going.

John Furrier

>> All right. Since you brought up Steve Jobs who lived in my town, Laurene and Steve, my friends that worked for him said he always has a question they would ask people in an interview process. So I'll ask you for Mill. What's the coolest thing you're working on right now?

Matt Rogers

>> Ooh. I think the coolest thing we're working on is this AI on the edge. It's really where the puck is going. I think about the amount of data center build out that's happening right now and the massive CapEx that we're building out. But if in five to 10 years really things are going to move to the edge, folks need to be start thinking about what their edge strategy really is and are we right sizing the CapEx for where the future is going to be.

John Furrier

>> And what's cool about the edge in your mind? What use cases, what unlocks, what's possible that we don't see today that comes down to the edge if that continues?

Matt Rogers

>> One is speed. Things at the edge are always going to be faster. And then the other is cost. The cost of building data centers both from a CapEx and an OpEx. The energy it requires to run. New silicon comes out every few years. That cost is enormous versus if you could build a special purpose AI engine at the edge that does everything that you need to do in that location, oh man, that's a much more efficient operation.

John Furrier

>> I think your venture, I think is a tell sign at least is my vision on this, is that what you're doing is one example of many that will come. If you think about our homes, and I don't know if you remember the days back in the 90s, grid computing. It was different context, but imagine having a grid of factory nodes. We have energy. New York's the biggest waster of energy here. I mean they could have enough energy aggregated from New York-

Matt Rogers

>> That's right.

John Furrier

>> If there was brains.

Matt Rogers

>> Yeah.

John Furrier

>> Because you're putting brains into your system. So imagine if I want to manage my boiler from a high-rise building, that's an IoT device.

Matt Rogers

>> That's right.

John Furrier

>> That's AI device.

Matt Rogers

>> It just kind of reminds me of the old folding at home.

John Furrier

>> Yeah.

Matt Rogers

>> Or the SETI at home days of yore where folks would offer their personal computers at home to work on big compute tasks. I think we're getting to the point where not just are our personal computers and our phones very, very high horsepower, but the rest of our edge devices are going to be there too.

John Furrier

>> Yeah.

Matt Rogers

>> We saw this coming at Nest and we're at the point now where-

John Furrier

>> In which way did you guys see it?

Matt Rogers

>> That more and more could be pushed to the edge and we started this effort called Thread to build more mesh networking to allow nodes to talk to each other and share intelligence. It is where the future is going. The cost of energy, the cost of data centers is really high. And I think about the average application, the average application of AI, the full build out of the data center is too expensive for those average applications.

John Furrier

>> How do you see the intelligence connecting into your commercial business? Because imagine that Whole Foods, they have a lot of locations in all the major metro areas, almost like in all these high bandwidth areas. The demographics in Whole Foods is pretty high, a lot of connectivity. What are you seeing that's going to be future headroom for you with the intelligence devices talking to each other, not just systems of record, but do you see, kind of like what's the vision of how this unfolds for Amazon and the world?

Matt Rogers

>> Look, I think to start with, if we could help solve their food-based problem, that alone is like a multi-hundred billion dollar problem to solve let alone thinking about as a whole, the waste problem.

John Furrier

>> Yeah.

Matt Rogers

>> The amount of waste we create as a society, massive. Actually just the value of wasted food is maybe twice the size of the waste industry. The value of the things we throw away is immense.

John Furrier

>> And what's the growth strategy now? Take us through where you are in the progress you got. Is this shipping product with the commercial side, is it iterating through? Take us through the progression.

Matt Rogers

>> Right, okay. So we're in market today with our residential product available at mill.com. People are buying it all the time every day. We've announced our commercial scale product and have started taking pre-orders for that and we're starting to deploy prototypes later this year.

John Furrier

>> Is there a threshold for customers on the orders? And you're prioritizing as obviously Amazon's huge so it's good work case for you.

Matt Rogers

>> I mean, we're starting with the most innovative customers and kind of a canonical customer per category. Think about Whole Foods as being a canonical grocery store. We're also looking at restaurants, stadiums, food service, universities, hospitals, innovators in their space.

John Furrier

>> Got it. All right, so what's on your agenda? What are you focused on now personally with the company, status, hiring, put a plug in.

Matt Rogers

>> Yeah, so always love meeting with customers, meeting with folks in the waste industry. I spent a lot of time in DC talking about the waste issue. Meeting with mayors around the country, talking about waste at a local level. This is one of these issues that pretty much everyone get behind. There's no pro-waste-

John Furrier

>> No negative.

Matt Rogers

>> People out there.

John Furrier

>> Yeah, you're not getting a lot of nos. More like how do you do it? What's the cost? Is it viable?

Matt Rogers

>> That's right. You got it.

John Furrier

>> And what's your answer to that when they say, okay, we love it. What's the cost? What's the blocker? Why isn't it going faster? Is there evolution?

Matt Rogers

>> I think at this point we've found that product market fit where the product has a clear return on investment and pays for itself. And once you find that, right, then there's no green premium, this thing will sell itself.

John Furrier

>> All right, so for the folks that are out there watching with your history, I love the little throwback there with Apple and Nest. For people who are trying to run their business right now, they're going through the strategy risk. That was last year.

Matt Rogers

>> Yes.

John Furrier

>> This whole AI. This is an execution and risk mode right now because it's pretty obvious where the world's going. So it's an AI or die. That's almost an extinction event. It's like the internet dying, killing a lot of other things. It just gets better.

Matt Rogers

>> Yes.

John Furrier

>> What advice would you give folks as they start thinking about their operations, their products, how to optimize their people, process, and technology for the future? And I mean, and obviously they're not going to lay off the entire company and productivity kicks in. That's an easy one. But what else should executives and leaders be thinking about as technology drives their business model transformation?

Matt Rogers

>> Be open to the change and be out there learning about new solutions. And if you don't have someone on your leadership team whose kind of your forward scout, you should probably have one who's doing it. We certainly do. And I think about the change that's happening in this space is a change that we've felt and seen before, but maybe not this fast.

John Furrier

>> The speed. So speed is key.

Matt Rogers

>> The speed is totally different. I think about the blockbuster to Netflix transition or the Blackberry to iPhone transition or the web to mobile transition. Those transitions took years. This one feels different. It feels faster.

John Furrier

>> And what's a forward scout look like?

Matt Rogers

>> Someone who's meeting with the AI platform companies who's aware of all the new AI applications that are being built. Things like Harvey that's automating a lot of legal software, things like that. What are the specific applications that could apply to your business and make you faster?

John Furrier

>> All right, Matt, thanks for coming on. I really appreciate it. Any predictions for this year just in the industry? Market's down a little bit today. It's up and down. The SaaS company's going up and down, AI infrastructures, people trying to figure that out. What's your take on the technology as the market?

Matt Rogers

>> I'm excited. I think this is going to be a really exciting year. I look forward to some really big tech IPOs later this year. I think that's something a lot of us have been waiting for, for a long time.

John Furrier

>> Yeah.

Matt Rogers

>> I think it's going to be exciting year.

John Furrier

>> Great to have you on. Congratulations.

Matt Rogers

>> Thank you.

John Furrier

>> Great story. Not only are they doing some really cool technical things, it's really at the center of a lot of societal problems. Food waste, waste is a big problem. The money's quantifiable. The business benefits are massive. But also from a climate change and earth standpoint, societal benefits are off the charts. These are kind of the missions. We'll see a lot more of these missions and the capitalism behind it, making it happen and funding it. Of course, theCUBE's doing its part to bring it to you. Thanks for watching.

Matt Rogers, Mill

search

John Furrier

>> Welcome back. I'm John Furrier with theCUBE. We are here in our NYSE studio. Of course, we have our Palo Alto studio connecting Silicon Valley and Wall Street technology is the market. All the stocks, all of our lives are infused with technology, the smartphone, cloud generation. Now the AI native generation is upon us. The AI infrastructure is expanding very, very fast. And again, we'll continue to tell that story of AI factories, AI factor to the edge. Of course the real world and digital world is coming together. Matt Rogers here, CEO of Mill, a company tackling a unique problem, but it's kind of a climate change opportunity, but it's also a very clever device. Also, the co-founder of Nest. Great to see you again. Thanks for coming back in.

Matt Rogers

>> Great to be back. Thanks for having me.

John Furrier

>> So I just love your pedigree, if I can say that word, Nest, Apple, you're tinkering, you're building, you're engineering your new venture. Give a quick update because you were in early. I know there's a lot of updates I want to get to quickly, but give a quick overview of what you guys are doing.

Matt Rogers

>> Yes. I'm building a company called Mill and we're building a waste technology company where we're combining AI and hardware to prevent waste and build a new waste system. We started with a product at home, a residential food recycler, which helps people manage food waste at home. We recently announced just a few months ago a new commercial scale version where we're deploying this first to Amazon and all their grocery stores and Whole Foods locations where we're also building an AI, a camera technology that's going to identify what food's being thrown away. It's a pretty holistic product that not just is helping manage waste but prevent waste altogether.

John Furrier

>> So check out that video on YouTube last time you were in. It was a home product, but it was very interesting because you're using the technology to do all that in this so there's not a lot of waste. Now the commercial business, like businesses who throw out a lot of stuff, restaurants, food, shopping malls, all that's getting wasted. You have technology. Explain the secret sauce to what you guys do.

Matt Rogers

>> Yes. So about half the food we waste comes from homes, the other half from businesses. The core of our technology is dehydration. So most of food is actually just water. So you think about those garbage trucks that are driving in our neighborhoods, 80% of food is water. So this is a lot of water being trucked around. So the core of our product is we take the water out of food and when you dehydrate it, it gets small, almost looks like coffee grounds. Doesn't smell, doesn't go bad. Also, you don't have take the trash out as often, but if you think that as a business, not just are you picking up the trash less frequently, but if you add AI, could you identify what food is being thrown away, when it's being thrown away and why? And then could you feed that back into your procurement systems, your menu planning to prevent food waste altogether and really save money?

John Furrier

>> I really liked the practical approach you took on building this company. One home, half the problems there. So tackle that first, get the product ready. It does what it does technically, solves that problem. Now you say, okay, let's take a bigger problem, commercial. Give us the updates. You got AI in there, you mentioned before we came on camera. Explain the advances of now a connected device could be better than a non-connected device.

Matt Rogers

>> It's pretty amazing what could be done today. So I started Nest about 15 years ago and for those, remember we had these Nest cam products, video doorbells, things like package detection where the UPS driver would drop off a brown package and you get a notification at home. Now we built that, we hand-tuned that model. It took like a year to build that. Dozens of engineers. Now with latest LLM technology, we could do this so much faster. What used to take years now takes weeks and the capacity to train these models, it's way faster. It's remarkable what could be done today.

John Furrier

>> So last time you were on last year, now this year. So what happened? You go back to the ranch, you go back to the team, AI is now hitting the table. We're starting to see those advances, acceleration. Jensen Huang calls it accelerated computing. You got super computing capability, you can buy it anywhere. GPU clouds, Amazon on a device. What did you guys do? What were the meetings like? We have to get AI in this thing. And well, take us through that.

Matt Rogers

>> There exists products and services out there today that try to help restaurants and grocery stores prevent waste. But usually they're additional work. There are things you add into your workflow and we were thinking is what if we could build it into what you do already? And today people take their foods, scraps, they put it in the trash can. Now what if we could intercept it just then, just when they're in the middle of throwing things away, not adding more time to their work? And this is exactly what AI is great at. And in this case, as you said, computing has gotten so good. We actually built the data center into the trash can.

John Furrier

>> Now that's an edge device. All right, so what's going on? So the computer vision, is there a computer vision? How does it detect?

Matt Rogers

>> That's right. We use a network of cameras plus sensors and we kind of combine all those together to get a really accurate picture of what's being thrown away. Because the vision technology is really good identifying what is being thrown away, but you need other sensors to determine how much they weigh and how much water is in them and that kind of stuff.

John Furrier

>> All right. Take us through the expansion into commercials. I find this fascinating because, and I'm curious how the AI is reporting the data. So you've got Amazon and Whole Foods, they have an operation, they have inventory. People can throw stuff away if they're done eating at Whole Foods, but they have the buffet there, they got tons of food, they got produce and they got tons of food that's perishable.

Matt Rogers

>> That's right.

John Furrier

>> Is that the waste? Take us through the problem and what you guys do to solve it.

Matt Rogers

>> So the core problem is food waste is several percentage points of revenue for these companies. And for those who know the grocery business, grocery business is not a high margin business. So a couple points of revenue is a big deal. So if we can go and identify what are the sources of those waste, when are they happening and agentically, what can we do to prevent that waste? Tying in automatically to their procurement systems or their menu planning systems. What times a day are these things happening and could we automatically make these inferences to prevent the waste from being happening?

John Furrier

>> So you take the edge device, which is the trash system with the data center and all of that. My mind's going crazy there. It's a rack of servers. And then you're connecting to essentially databases which are traditionally siloed workflows.

Matt Rogers

>> That's right.

John Furrier

>> You're tying that together. Is it agentic? Are you building in coordination, delegation? Take us through some of the things going on.

Matt Rogers

>> Exactly. This is an ideal use case for AI. This is actually exactly what agents are for. The kind of things of connecting these dots that are really hard to do manually and in the spirit of what's happening in the software world today, this is one where if we did it the old way, if we built custom SaaS solutions for every vertical, whether it's a custom software solution for quick service restaurants versus grocery stores, we would need thousands of engineers to do that. And this is kind of the ideal use case for agentic AI.

John Furrier

>> So I have to ask you Matt on, while you're here, since you're an expert. We have a mixture of expert series. So we'll kind of throw that in too. Yeah. The whole SaaS getting killed by AI conversation. I interviewed OpenTable last week, they were here and they been around for 20 plus years. They got tons of venues. Took a lot of work to get that supply chain down. They got the demand side with the user base. They're a SaaS app, they're like the textbook cloud app. I mean you get a restaurant order. I don't see them going away. So I mean, so where do you see the line between a competitive moat, either technically, system of record or data versus say the interface changes where things are being embedded in like you're doing with the waste systems or prompts or now the interface? What's your take on how you see, I'm sure you get this all the time, is SaaS getting killed by AI? What's your-

Matt Rogers

>> John, you know I've been doing this a long time.

John Furrier

>> Yeah.

Matt Rogers

>> We've seen the arcs of technology and think back to 1999 when Cisco is the most valuable company in the world. Right. The rise and fall of technology companies is one that's tale as old as last several decades. I think about companies like Nokia and RIM and what happened with the iPhone. The question with these SaaS companies is now the barrier to entry has gone down and custom solutions can be built much faster. How are they evolving their businesses? And one of the things that Steve Jobs taught us at Apple is you got to think about not just your competitive moats, but can you go after your own businesses?

John Furrier

>> You mean compete on value.

Matt Rogers

>> Compete on value. So-

John Furrier

>> And cannibalize your own business with better product.

Matt Rogers

>> You got it. So we were worried about this. This is kind of the genesis of the iPhone. We were worried. Hey, smartphones are coming. They're going to eat our iPod business, which was most of Apple's revenue at the time. So part of the genesis of the iPhone is kind of defending the fort on the iPod business. Now I think about it, if I was a software CEO right now, I'd be thinking about that. What are my competitive moats? Where's the puck going? And then start building solutions to where the puck is going.

John Furrier

>> And so that is classic business strategy product. Product market fit. Question isn't are we going to be around, the question is product market. That's what you're saying.

Matt Rogers

>> That's right. In this case the market is shifting and product is shifting at the same time. The competitive moat that a lot of SaaS companies had of it would take tens of thousands of engineers to build this product. That barrier has gone down.

John Furrier

>> Yeah.

Matt Rogers

>> So they have to think about their moats differently.

John Furrier

>> And that is the interface. All right, so back to where you're going with this because you're now again, one of my favorite topics that I'm working on is hyper-converged edge. You are a working example of a device perfectly positioned for being an edge device. You got compute in the device itself or the system. It's a system, it's not like a device, it's not a tool.

Matt Rogers

>> That's right.

John Furrier

>> You're building essentially an OS within the system.

Matt Rogers

>> Yes.

John Furrier

>> It's going to connect in. Eventually the model's going to want to come down to your device and make you smarter. How do you see that going on your vision? Because as the world gets smarter, intelligent wise, you're just another node in the network, in this case multiple nodes. Amazon has all over . There's their network. Then there's the external network coming together. How do you see that architecture emerging?

Matt Rogers

>> Yeah, I think actually even back to my time at Nest, my co-founder Tony Fadell and I envisioned this even back then. And at the time technology was just not good enough. We actually wanted to be able to build all the computer vision into the Nest cam itself and we were just not there yet. But we're at the point now where I think it can be done. And I think from a privacy standpoint, but also just like compute resources and energy perspective, it's much better to do these things locally. The more local the better and being able to just download a new model or take the inferences and learnings from all of the nodes of the network and build new updates to the model and push that back down to the edge. I think this is the way of the future. I think this is where things are going.

John Furrier

>> All right. Since you brought up Steve Jobs who lived in my town, Laurene and Steve, my friends that worked for him said he always has a question they would ask people in an interview process. So I'll ask you for Mill. What's the coolest thing you're working on right now?

Matt Rogers

>> Ooh. I think the coolest thing we're working on is this AI on the edge. It's really where the puck is going. I think about the amount of data center build out that's happening right now and the massive CapEx that we're building out. But if in five to 10 years really things are going to move to the edge, folks need to be start thinking about what their edge strategy really is and are we right sizing the CapEx for where the future is going to be.

John Furrier

>> And what's cool about the edge in your mind? What use cases, what unlocks, what's possible that we don't see today that comes down to the edge if that continues?

Matt Rogers

>> One is speed. Things at the edge are always going to be faster. And then the other is cost. The cost of building data centers both from a CapEx and an OpEx. The energy it requires to run. New silicon comes out every few years. That cost is enormous versus if you could build a special purpose AI engine at the edge that does everything that you need to do in that location, oh man, that's a much more efficient operation.

John Furrier

>> I think your venture, I think is a tell sign at least is my vision on this, is that what you're doing is one example of many that will come. If you think about our homes, and I don't know if you remember the days back in the 90s, grid computing. It was different context, but imagine having a grid of factory nodes. We have energy. New York's the biggest waster of energy here. I mean they could have enough energy aggregated from New York-

Matt Rogers

>> That's right.

John Furrier

>> If there was brains.

Matt Rogers

>> Yeah.

John Furrier

>> Because you're putting brains into your system. So imagine if I want to manage my boiler from a high-rise building, that's an IoT device.

Matt Rogers

>> That's right.

John Furrier

>> That's AI device.

Matt Rogers

>> It just kind of reminds me of the old folding at home.

John Furrier

>> Yeah.

Matt Rogers

>> Or the SETI at home days of yore where folks would offer their personal computers at home to work on big compute tasks. I think we're getting to the point where not just are our personal computers and our phones very, very high horsepower, but the rest of our edge devices are going to be there too.

John Furrier

>> Yeah.

Matt Rogers

>> We saw this coming at Nest and we're at the point now where-

John Furrier

>> In which way did you guys see it?

Matt Rogers

>> That more and more could be pushed to the edge and we started this effort called Thread to build more mesh networking to allow nodes to talk to each other and share intelligence. It is where the future is going. The cost of energy, the cost of data centers is really high. And I think about the average application, the average application of AI, the full build out of the data center is too expensive for those average applications.

John Furrier

>> How do you see the intelligence connecting into your commercial business? Because imagine that Whole Foods, they have a lot of locations in all the major metro areas, almost like in all these high bandwidth areas. The demographics in Whole Foods is pretty high, a lot of connectivity. What are you seeing that's going to be future headroom for you with the intelligence devices talking to each other, not just systems of record, but do you see, kind of like what's the vision of how this unfolds for Amazon and the world?

Matt Rogers

>> Look, I think to start with, if we could help solve their food-based problem, that alone is like a multi-hundred billion dollar problem to solve let alone thinking about as a whole, the waste problem.

John Furrier

>> Yeah.

Matt Rogers

>> The amount of waste we create as a society, massive. Actually just the value of wasted food is maybe twice the size of the waste industry. The value of the things we throw away is immense.

John Furrier

>> And what's the growth strategy now? Take us through where you are in the progress you got. Is this shipping product with the commercial side, is it iterating through? Take us through the progression.

Matt Rogers

>> Right, okay. So we're in market today with our residential product available at mill.com. People are buying it all the time every day. We've announced our commercial scale product and have started taking pre-orders for that and we're starting to deploy prototypes later this year.

John Furrier

>> Is there a threshold for customers on the orders? And you're prioritizing as obviously Amazon's huge so it's good work case for you.

Matt Rogers

>> I mean, we're starting with the most innovative customers and kind of a canonical customer per category. Think about Whole Foods as being a canonical grocery store. We're also looking at restaurants, stadiums, food service, universities, hospitals, innovators in their space.

John Furrier

>> Got it. All right, so what's on your agenda? What are you focused on now personally with the company, status, hiring, put a plug in.

Matt Rogers

>> Yeah, so always love meeting with customers, meeting with folks in the waste industry. I spent a lot of time in DC talking about the waste issue. Meeting with mayors around the country, talking about waste at a local level. This is one of these issues that pretty much everyone get behind. There's no pro-waste-

John Furrier

>> No negative.

Matt Rogers

>> People out there.

John Furrier

>> Yeah, you're not getting a lot of nos. More like how do you do it? What's the cost? Is it viable?

Matt Rogers

>> That's right. You got it.

John Furrier

>> And what's your answer to that when they say, okay, we love it. What's the cost? What's the blocker? Why isn't it going faster? Is there evolution?

Matt Rogers

>> I think at this point we've found that product market fit where the product has a clear return on investment and pays for itself. And once you find that, right, then there's no green premium, this thing will sell itself.

John Furrier

>> All right, so for the folks that are out there watching with your history, I love the little throwback there with Apple and Nest. For people who are trying to run their business right now, they're going through the strategy risk. That was last year.

Matt Rogers

>> Yes.

John Furrier

>> This whole AI. This is an execution and risk mode right now because it's pretty obvious where the world's going. So it's an AI or die. That's almost an extinction event. It's like the internet dying, killing a lot of other things. It just gets better.

Matt Rogers

>> Yes.

John Furrier

>> What advice would you give folks as they start thinking about their operations, their products, how to optimize their people, process, and technology for the future? And I mean, and obviously they're not going to lay off the entire company and productivity kicks in. That's an easy one. But what else should executives and leaders be thinking about as technology drives their business model transformation?

Matt Rogers

>> Be open to the change and be out there learning about new solutions. And if you don't have someone on your leadership team whose kind of your forward scout, you should probably have one who's doing it. We certainly do. And I think about the change that's happening in this space is a change that we've felt and seen before, but maybe not this fast.

John Furrier

>> The speed. So speed is key.

Matt Rogers

>> The speed is totally different. I think about the blockbuster to Netflix transition or the Blackberry to iPhone transition or the web to mobile transition. Those transitions took years. This one feels different. It feels faster.

John Furrier

>> And what's a forward scout look like?

Matt Rogers

>> Someone who's meeting with the AI platform companies who's aware of all the new AI applications that are being built. Things like Harvey that's automating a lot of legal software, things like that. What are the specific applications that could apply to your business and make you faster?

John Furrier

>> All right, Matt, thanks for coming on. I really appreciate it. Any predictions for this year just in the industry? Market's down a little bit today. It's up and down. The SaaS company's going up and down, AI infrastructures, people trying to figure that out. What's your take on the technology as the market?

Matt Rogers

>> I'm excited. I think this is going to be a really exciting year. I look forward to some really big tech IPOs later this year. I think that's something a lot of us have been waiting for, for a long time.

John Furrier

>> Yeah.

Matt Rogers

>> I think it's going to be exciting year.

John Furrier

>> Great to have you on. Congratulations.

Matt Rogers

>> Thank you.

John Furrier

>> Great story. Not only are they doing some really cool technical things, it's really at the center of a lot of societal problems. Food waste, waste is a big problem. The money's quantifiable. The business benefits are massive. But also from a climate change and earth standpoint, societal benefits are off the charts. These are kind of the missions. We'll see a lot more of these missions and the capitalism behind it, making it happen and funding it. Of course, theCUBE's doing its part to bring it to you. Thanks for watching.