Repeat founder Agustin Lopez Marquez, ex-nference, brings his AI savvy to food delivery with his new startup. Built on CloudKitchens’ infrastructure, Mowgli addresses pain points for diners, sous-chefs and drivers. Loved my chat with this brilliant and practical founder. Sample Mowgli’s homey meals from exotic places at https://khipi.com/ (delivers only to Boston & Cambridge for now)
Highlights:
Sal Daher Introduces Agustin Lopez Marquez
"... my co-founder, who happens to be my wife, Brinda, she was the one that started the company and the first product about a couple of years back..."
"... The idea here is that these are meals for explorers, for people who like to venture out, travel. This is not pizza or your usual weekend fare for the kids, and so forth..."
"... it's an infrastructure play that has happened in the last three to five years. It's modular kitchens that are just the kitchens that are meant for delivery-only companies..."
"... You're going to become like the McDonald's of adventurous eating..."
"... we already have more than 10 drivers that consistently, every Tuesday, they choose to work for Khipi as opposed to working for any other food delivery company..."
"... I understand you have 400 customers ordering right now. 90 of those are super dedicated and they order every week. They give you 80% of your business..."
ANGEL INVEST BOSTON IS SPONSORED BY:
Transcript of “AI for Discovery Dining”
Guest: Agustin Lopez Marquez
Check Out Sal’s Substack on Aging, Fitness & Connectedness
Sal Daher: Hey, this is Sal Daher. I’m delighted you found the Angel Invest Boston podcast in which I interview people who know a lot about building technology startups.
I now have a Substack focusing on exercise, healthy eating and human connectedness, and occasionally, startup investing. It’s called Aging Healthy & Connected.
If you are interested in my perspectives on these subjects look for Sal Daher on Substack.com. Daher is spelled Delta Alfa Hotel Echo Romeo.
Enjoy the podcast.
Sal Daher Introduces Agustin Lopez Marquez
Welcome to Angel Invest Boston, conversations with Boston's most interesting angels and founders. I am Sal Daher, an angel investor who's really, really fascinated to discover how to better build technology companies. Today, we are very lucky to have a phenomenal founder. Welcome, Agustin Lopez Marquez.
Agustin Lopez Marquez: Hi, Sal. Very nice to be here.
Sal Daher: Now, Agustin is a guy who spans disciplines. He was involved in the founding of SQZ Biotech, which is very much a wet lab biotech company, chemical engineer by training. Chemical engineers are the brainiest of engineers. Aero and astro guys think they're the smartest, but my brother-in-law, who's a very brilliant patent attorney, Peter Fasse, says chemical engineers are the smartest engineers. They know the math, they know the science, they're really practical, but they can do theory.
Anyway, so Agustin went off after SQZ and he went to work for nference, which is a company that's applying AI to create opportunities for researchers to discover new things. It's basically using databases and trying to figure those out. This is way, way before all the big advances that we've had with generative AI and so forth. He's been working in this field for many, many years.
Agustin now has started a company, an AI company, but a totally different domain. The thing with AI is, AI is an all-purpose tool. He is applying his orientation towards scientific discovery process and AI to a very, very interesting field, which is dining. Anyway, take it away, Agustin, and explain to us what Mowgli does for the consumer at present.
"... my co-founder, who happens to be my wife, Brinda, she was the one that started the company and the first product about a couple of years back..."
Agustin Lopez Marquez: Perfect. I'm excited to tell you, Sal. Mowgli is a food delivery company. We cook and we deliver food to customers. This food is amazing, and it's unique. It's not your common tacos, and burgers, and pizza. It's the type of things that you wouldn't find easily. Not only that, but we deliver it to customers at a price that is a lot cheaper than current food delivery companies like DoorDash and others.
You might ask, what's the catch? How can we deliver amazing differentiated food at a price that is a lot better than current alternatives? The catch is that customers order ahead of time. We take orders anywhere from a couple of days before, up until a day before they get the food. That, Sal, that one day ahead of ordering enables the behind-the-scenes of the company, which is the operations, that allows us to deliver that cost structure that we can pass to customers.
That operations is completely tech-enabled, so we use anything from-- Sal, you were spot on on the background because we use-- When I say tech, I'm referring to anywhere from chemical engineering, so some principles on how we do the cooking, all the way to AI and robotics across the board, from cooking to delivering.
It's a great experience for me personally. I've always loved food, but I've never been in the food industry. Actually, my co-founder, who happens to be my wife, Brinda, she was the one that started the company and the first product about a couple of years back. It's been a year since the two of us decided to scale it and create Mowgli as a company that will scale that initial product that she designed and started about two years ago.
"... The idea here is that these are meals for explorers, for people who like to venture out, travel. This is not pizza or your usual weekend fare for the kids, and so forth..."
Sal Daher: Excellent. The idea here is that these are meals for explorers, for people who like to venture out, travel. This is not pizza or your usual weekend fare for the kids, and so forth.
Agustin Lopez Marquez: Yes.
Sal Daher: These are dishes from different places in the world that you would not be able to get delivered, but that are-- I can tell you the secret sauce, so to speak, because I love to eat, by the way. The secret sauce is that the team at Mowgli, they're mining folk recipes. They're using well-established recipes from folk cuisine which are tried and true. It's a recipe from some region in India, a recipe from some region in South America, from some other place which has been perfected for centuries to be unbelievably tasty. However, you cannot do that at scale. You cannot deliver these tasty recipes at scale unless you're a chemical engineer, and you understand things like heat transfer. How do you cook something optimally? Is it in large trays, smaller trays? Things of that nature. It's really fascinating to me that Agustin Lopez Marquez is taking all this knowledge, all this experience, and his wife's good sense and inventiveness. She's really the one that started this-
Agustin Lopez Marquez: Yes.
Sal Daher: -if I remember the story correctly. She started experimenting with this and understanding of human beings. What we're doing is we are sampling the most tasty recipes from all the massive variety of cuisines that exist in the world.
I'll just give you an idea. I remember, back in the '90s, reading an article by the chief chef of Singapore Airlines. He said they estimate that Chinese cuisine is the cuisine with the most variety of recipes. There are probably 80,000 recipes. Arabic cuisine is 30-something thousand, 40,000. Then there are other cuisines that have fewer recipes. The point here is that we are overwhelmed with choices, and your wife's idea of creating this restricted poulet that has only gorgeous, delicious things in different-- I understand it's two choices per week.
Agustin Lopez Marquez: Yes.
Sal Daher: You choose one-
Agustin Lopez Marquez: That's correct.
Sal Daher: -or the other. I understand you have numbers that are through the roof in terms of people signing up, again, for this wonderful experience.
Agustin Lopez Marquez: That's correct, Sal. The company's called Mowgli, but our first product and flagship product is called Khipi. It's spelled K-H-I-P-I.
Sal Daher: If my daughter and my son-in-law wants to order from there, where would they go?
Agustin Lopez Marquez: They go to khipi.com.
Sal Daher: K-H-I-P-I.com?
Agustin Lopez Marquez: .com, yes. Every Thursdays, like today, actually, about an hour ago, we released the menu for the week. If people want to go right now, it's a good time.
Today, for example, we are offering keema rolls from India. Actually, this is from West Bengal, which is where my wife, Brinda, was born. Also, yassa poulet, which is our chicken from Senegal, which is a national dish in Senegal.
The way it works, Sal, is, every week, there's two dishes. It's for very simple decision-making within a week. Every week, we change the countries. We have done already almost 70 countries, 160 unique dishes in the last two and a half years. It's a perfect balance for the right customer profile for that adventurous eater. You get the diversity of trying new things that you wouldn't try otherwise, but within a week, you get the simplest decision-making of not just food, Sal, but our aspiration is that there's nothing easier than ordering Khipi. It's like you have only two options and that's it. It's fantastic for those adventurous eaters.
It's fantastic because it's busy people that love to learn about cultures, but they're also busy, of course, so they don't want to think too much. That's why we're seeing incredible retention, especially for that adventurous group of consumers.
Sal Daher: Wow. This is really very appealing. Listeners should check out Khipi, K-H-I-P-I.com, and look at the recipes every week. Wow, $15.
Agustin Lopez Marquez: Yes.
Sal Daher: It's not going to come tonight. It's not going to come Friday. It's next week, right? By Monday-
Agustin Lopez Marquez: Exactly.
Sal Daher: -people would choose what they want, and then gets fulfilled when?
Agustin Lopez Marquez: On Tuesday. You can order anytime from today, when we release the menu, anytime until Monday noon, and then you get the food delivered on Tuesday.
The food is ready, Sal, to be eaten after you put it in the microwave. It's just a matter of heating up, and then you're ready to eat it. It's very, very simple.
Sal Daher: This is amazing. You can choose a single, or you can choose a double, or a family portion. $40 for a family, double's $28. This is fantastic. Now, this is the visible part of it. Let's step back and see what is it that's going on to fulfill this. I understand that you're using CloudKitchens?
"... it's an infrastructure play that has happened in the last three to five years. It's modular kitchens that are just the kitchens that are meant for delivery-only companies..."
Agustin Lopez Marquez: That's correct, Sal. Actually, one of the reasons I love this idea from the beginning was that the customer side is very straightforward. If you're an adventurous eater, you can just explore new cuisines, and the decision-making is easy, and the price is really cheap. It's an ideal business idea where that part is very clear, but the operations is what's really unique.
Let's start from why it's the right time to do this. Closed kitchens as you were saying, CloudKitchens in particular, it's an infrastructure play that has happened in the last three to five years. It's modular kitchens that are just the kitchens that are meant for delivery-only companies. The thinking was that as the world moves more and more to people ordering food online to be delivered at home, it makes little sense for the companies that are cooking, having to have a high-cost real estate where people come to dine, as opposed to just having the tiny kitchen in an inexpensive area where you can just cook and then deliver the food at home.
The largest company in the space is called CloudKitchens and it was founded by Travis Kalanick, which people would know as the founder of Uber. It's really exciting, Sal, because from an infrastructure perspective, when I saw those kitchens, I thought it was the dream of a scaler because they already have 80 locations, they have tons of available kitchens.
It's very easy for us once we have the scalable formula here in Boston, to move to Seattle, to move to San Francisco, to move to Colorado because they already built the kitchens for us.
Sal Daher: Well, you mean not to move, but to expand to Seattle, expand to Colorado.
Agustin Lopez Marquez: That's right.
"... You're going to become like the McDonald's of adventurous eating..."
Sal Daher: Replicate the formula. You're going to become like the McDonald's of adventurous eating.
Agustin Lopez Marquez: That's correct.
Sal Daher: It's not going to be homogenized, pasteurized, it's going to be high-quality food. Let's step back a little bit here and remember, this is not fast food. This is not food prepared in a pizza joint. These are foods prepared in a professional kitchen by professional chefs. You have three sous chefs that you hire to fulfill the orders. These are high-quality orders. The recipes have been curated with having this delivery model in mind, and you're also thinking very hard about creating a good quality of life for the chefs that are cooking and a good experience for your delivery staff.
Agustin Lopez Marquez: That's correct.
Sal Daher: There are three constituencies here. It sounds like a complicated business, but because it's a food delivery business, it entails all these things.
Right now, you're really boning up on understanding customer preference, on getting, figuring out recipes that people really, really want to eat.
Agustin Lopez Marquez: That's correct.
Sal Daher: Helping them choose them and so forth.
Agustin Lopez Marquez: That's correct.
Sal Daher: The potential of the company is massive because you have all these other three components, and then I understand-- I don't want to speak out of turn here, but this is in your eventual roadmap, is automation of a great deal of the process, so that-
Agustin Lopez Marquez: That's correct.
Sal Daher: -not only are you improving the efficiency of the process in terms of helping people decide better, helping discover really exciting recipes, you're also just making the whole process less expensive, making it affordable for you to be eating adventurously.
Agustin Lopez Marquez: That is correct. I'll start with the McDonald's comment because we use McDonald's as an inspiration. I think McDonald's, if you think about it from a branding perspective, was taking the American food, burgers, globally. It's been very interesting that over the last decades, there's been all this globalization that has happened, but there's not a truly global food company.
Sal Daher: Coming the other way. That's awesome.
Agustin Lopez Marquez: Coming the other way. We think it's about time. It's been decades long that this should have existed.
Now in terms of the stakeholders, you're spot on. There's customers, there's cooks, and there's delivery drivers. What is fascinating for us is that in the last year, we have really nailed what is it that each of them want to be able for this to work at scale. We have talked about customers. Let's talk about sous chefs and cooks for now.
If you look at a traditional restaurant, typically, cooks and sous chef would work late nights, weekends. It's what we call a real-time model because people come and they order within like-- You want your food delivered at the moment in time, or if you're delivering online, within 30 to 60 minutes. It's this very high intensity, very stressful situation, especially around lunch and dinner time. Imagine if you're a sous chef and you have a family, you have kids. It's not great having to work nights and weekends. It's not the best life-work balance.
One of the really nice things about taking the orders ahead is that you can plan much better when you do the prepping, the cooking, the plating. Instead of having to do these things real-time, you can schedule them around normal times of operations, so you don't have to work nights, you don't have to work weekends. That's been the first, like the life component of why people are like, "Wow, this is really interesting. I could have a normal life. That's wonderful."
The second thing, Sal, is that for sous chefs and cooks is if you are in the same restaurant, over time you can get bored of doing the same things over and over, over years in some cases. It's really nice that, for them, they get exposed to different techniques and different cuisines every week, but that you do it in a controlled manner where you know ahead of time how many of those things you're doing and that you're doing only two of those things in a week.
Over time, you're getting trained across techniques and cuisines, but you're doing it in a very calm manner every week, doing two at a time.
Sal Daher: Right. It's not the MIT experience that you're drinking from a fire hose.
Agustin Lopez Marquez: Correct.
Sal Daher: It is more like you're able to sip because you're an artist. You're not just some kid scrambling to cram as much knowledge in your head as you can. You're an artist, you're learning different techniques which you might make variations on and so forth. You need to have a little time to absorb it, to process it, to become really, really proficient at it, highly proficient.
Agustin Lopez Marquez: That's correct.
Sal Daher: Then practice it from time to time, not be doing it every day. You're not a machine, you're a human being who will develop this whole repertoire of cooking techniques over time.
Basically, Khipi would be providing a tour of different types of cooking, because right now, just the two dishes you have here, one, the yassa poulet is like a stew. The other dish is basically like tacos, but very different flavors and so forth. It's different formats, which create complexity for you. Please talk about the delivery drivers, and then let's talk about automation.
Agustin Lopez Marquez: That's correct. Let's talk about the drivers. If you want to be a food delivery company, you have to deliver the food. If you look at the current delivery paradigm, it's dominated by DoorDash especially, but also Uber Eats and GrubHub to a lesser extent. When you look at that paradigm, the customers order to get the food within 30 to 60 minutes, but it comes with an 80% premium of what you would pay at the store. From a customer perspective, food delivery companies have been a great thing.
If you look at the market, it's $100 billion revenues annually. It's been, by many metrics, very successful, but if you ask people why don't they do delivery more often, they'll tell you the first reason is that it's very expensive. They feel it's a lot cheaper to cook at home. That's why the majority of people still cook at home, Sal, even though most people say they hate cooking at home. That's one of the biggest reasons why we're doing this, is how can we do food delivery, keeping all the good things about current food delivery, but make it a lot more affordable?
"... we already have more than 10 drivers that consistently, every Tuesday, they choose to work for Khipi as opposed to working for any other food delivery company..."
In terms of how it works for us, so we decided to tackle delivery, meaning that we have our own delivery software. The financial explanation of that is that the current model of DoorDash and others is that they take 30% of your revenues, top line for restaurants. We thought, "Is there a way to make a delivery app in our paradigm of one day ahead where we can make it ourselves? Would it cost us a lot less than 30% of the revenues?" The answer is that, yes, we have already built an app that our drivers are using. Sal, to talk about the driver side, which is the third stakeholder group, we already have more than 10 drivers that consistently, every Tuesday, they choose to work for Khipi as opposed to working for any other food delivery company.
Also, Sal, not only that, as opposed to working for Uber, or Lyft, or Amazon. You might ask, why are they choosing Khipi, if they could have all these other options, is that we also tackle something very important for the drivers, which is, if you look at the current food delivery paradigm, you send a driver to a restaurant, and then you send that driver to a customer. Then you keep repeating that, so you go to different restaurants, and different customers.
In our case, what we do is we give drivers predictability, because we tell them, "Hey, you take 20 orders of 20 customers at one time." That, Sal, is already reducing by half the number of stops, because you don't have to-- You only have one restaurant as opposed to 20, but not only that, the route is already optimized. It's not even half the distance, it's actually lower than that because you're not sending somebody from Lexington all the way to downtown Boston and then back to Arlington. Everything is optimized from that perspective.
Drivers are also loving, Sal, something that is-- It's a good segue for you if you don't mind to talk about the AI components of what we do.
Sal Daher: Sure.
Agustin Lopez Marquez: It's not only that you do less stops and it's predictable, but it's also if you think about-- As an example, I'm a big DoorDash user. I particularly love to use it for-- This is a shout-out for Tatte the bakery. I also use it for flour, so I don't discriminate. I love food delivery because of the convenience that it brings. I don't love the price many times, of course, but one of the little things of the experience that really matters to drivers and customers is that I, for example, I live in an apartment building. Every single time that they're going to deliver the food to me, they call me and they ask me, "Agustin, do you want it in the front?"
Sal Daher: "Where do I drop it off?" Yes.
Agustin Lopez Marquez: "Do you want it in the back? Do you want it in the gym? Do you want it in your door?" Every single time, hundreds of times, I've had the same chat, and I tell them I want it right in the front of the building. What would be the ideal experience for drivers and customers, Sal? For the customer, it's like, if I told you the first time it's in the front, don't ever ask me again. It's in the front. It's okay for me to tell you one time, maybe twice, but I don't need to tell you 200 times. It's unnecessary.
For the drivers, it's the same thing because it's like, they call you because they truly don't know. How do you solve this problem? First, there's a technology part, Sal, that is not AI, which is-- It's as simple as, the first time you do it, capture the instructions, capture the picture, and put it in front of the drivers in the app. So when the drivers go, they know, "Agustin likes it in the front, and this is a picture." With that, Sal, no AI, nothing.
Sal Daher: It's pretty simple.
Agustin Lopez Marquez: Where the AI comes in is that people still will make mistakes. Humans sometimes, we are-- You might get confused or whatnot. Sometimes as people make mistakes, what we do is we use AI to have the reference picture be compared against the picture that is taken by the driver. Real time you can compare, "Is this the same house and is this the same spot?"
It's fascinating, Sal, because in the past, back at nference, the previous company I was at, we used to do AI in imaging for things like pathology images, for things like radiology images. This is exactly the same underlying technology to your point at the beginning of the call, but its used in a very different context, which is, is this image similar enough to this other image?
Sal Daher: That's the point I was making about AI. It's an all-purpose, so machine vision, it can tell you about the porch, or it can tell you about imaging of some x-ray or whatever. To the AI, it's all patterns.
Agustin Lopez Marquez: Exactly, Sal. Those are the things I do. It's nice for each one of these stakeholders. There's a business model component of-- The one day ahead allows us to batch the orders such that you have fewer staff, you have an optimized route, you have predictable income for the drivers, but also the technology allows you to have a better experience for both the drivers and the customers. It's a part that makes us very differentiated.
I would also say, Sal, there's a fourth group I would put as the stakeholders, which is technology people. How do we build these AI technologies in-house, right? We need to hire people that could be working at SpaceX, or they could be working at nference, or they could be working at Google. Why are they going to be working at Mowgli? Well, it's been really interesting because I saw it personally, Sal. I would say I'm one of the fourth group. When I joined, and I always joke, but this is the mom test. I was talking to my mom and I was like, "Mom, I'm going to do this food company with Brinda." Initially, she was like, "Are you sure you're going to go from trying to cure cancer to cooking arepas?"
Sal Daher: Moms never understand that.
Agustin Lopez Marquez: It was a hard sell, like, "I thought you went to MIT to work in biotech?"
Sal Daher: I know. They never understand that.
Agustin Lopez Marquez: What's really interesting is that I realized that, as always, whenever you go off the path, and everybody walks, you realize that there's opportunities. The first thing I realized was-- First is, food historically, Sal, has been-- In terms of being a sexy space, people love great restaurants, right? Even chefs have traditionally be considered like, "Wow, that's a really cool job," even though it's a really hard job.
Culturally, people have seen the food industry to be cool, but nobody has actually, traditionally, wanted to especially be in the cooking side of that. The way we saw it was more, once we started actually cooking, and plating, and delivering, was understanding, are there enough intellectually difficult and interesting problems that people technology-wise would find, "I really want to solve that problem. That's really cool."?
Sal Daher: Yes. I see where you're heading. Some kid wants to be an astronaut in space, so they go to SpaceX. There are a lot of kids who love food and who want to be in SpaceX for food.
Agustin Lopez Marquez: That's correct.
Sal Daher: It goes along the direction of-- Everybody knows if you want to get grad students to an event, you got to have food.
Agustin Lopez Marquez: Typically, the dogma is, if you're a technology company in the food space, you don't want to touch the food because of the complexity. If you look at DoorDash and Uber Eats and others, we have taken that and actually turned it into an asset because we are like, "Look, if you love food and you love technology, here you're going to get to even smell it, touch it, taste it, plate it, deliver it." It's really exciting, which might be a good segue to talk about robotics that you brought up early on.
Sal, one of the most interesting things for me as still fairly a newbie to this space, it's been a year, we've delivered tens of thousands of plates. Maybe I was a real newbie about a year ago, but nowadays, we have a lot more experience. Still, it's been nice to see the space with fresh eyes.
One of the things that has been so clear is that in terms of labor, there is one step in the process that is very time-consuming, that is very repetitive, which is the plating of the dish, which if you're going to get the yassa poulet from this week, it means, Sal, that we need to put a chicken thigh. We have to put some curry on top. We have to put some rice on the side, but you have to do it one plate at a time. When you're doing hundreds of plates, this is a very time-consuming and very repetitive labor.
When I was seeing that problem myself week to week, I was like, "Wow. This is really a great space for robotics." Then when you ask the question, "Why hasn't robotics really solved that problem yet in the food world?" you realize that actually from a robotics perspective, it's actually a very complicated problem. The typical robotics problem is one where you have constant form factors. Imagine, I don't know, if you're in car manufacturing, you have the tires, the chassis. All of that is similar. It's similar form factors. Here, a chicken thigh from another chicken thigh is completely different.
It's very interesting that in the last couple of years, robotics have really exploded in terms of how amazing these robots are, but also how much more inexpensive they are compared to what they used to be. It's a really good time to embrace robotics. We have a team of people coming onboard that are robotics experts coming from MIT and other places that we have reason to believe that it's the right time to solve them.
"... I understand you have 400 customers ordering right now. 90 of those are super dedicated and they order every week. They give you 80% of your business..."
Sal Daher: Wow. This is really tremendous because it's interesting that you're starting off in chewable bites, if you don't mind the pun here. You started out with this restricted menu and this mode of weekly ordering as a very good experimental test bed for the concept. I understand you have 400 customers ordering right now. 90 of those are super dedicated and they order every week. They give you 80% of your business.
Agustin Lopez Marquez: That's correct.
Sal Daher: The other people are occasional, which is typical. I suspect when you go from 400 to 4,000 customers, you're probably going to be seeing pretty much that same power law distribution. Then you're going to be able to start looking at optimizing other aspects of this. As you grow, as you scale, you can then think about getting into fine-tuning things, such as the plating.
Agustin Lopez Marquez: That's correct. I think one of the growth strategies is that, you are correct that 80% of our revenues come from these power customers that are the type, Sal, that come every week basically, every two weeks at the most, but these people are extremely regular. Something that is very exciting, Sal, is not only the number of customers, more than 400 now, but it's also the spend per customer. These power customers are spending in Khipi more than they spend in Amazon.
An average customer of Amazon as an example, they're spending more money, more than DoorDash's average customer. It's very exciting seeing that for these power customers, Khipi is godsend to them. It's also really interesting for us that we know exactly who these power customers are from the perspective of characteristics that we find them. It's not surprising that-- If there's one question, Sal, that you could ask to know if somebody is likely a power customer or not is how many countries they have traveled. It's people that love traveling, so 90% of our power customers have traveled to 10 or more countries.
If you look at the US population, that's only 11% of the population. For that 11% of the population, Khipi will be very, very attractive. The total addressable market is very large because 11% of the US population is substantial.
Sal Daher: It's huge number of people. Yes.
Agustin Lopez Marquez: It's a huge number of people, but then the way to grow, Sal, is that over time, the same operations behind the scenes is something that we could use to go after other markets of non-adventurous people and more common cuisines. It makes perfect sense for us to start in a market that is very niche, if you will, where we are facing little to no competition because the offering is so differentiated. At the right time, where we're big enough, we can flip it and say-- we could start having offerings that go after the larger market of more common cuisines like the ones that we all know.
Sal Daher: Right. This is such a happy coincidence. This morning I interviewed Todd Zion, the founder of SmartCells, of Akston Bio, who's a very well-known founder in biotech, very successful founder. One of the things he said is that they tried to create a protein-based COVID vaccine in the middle of COVID. They have it on the shelf. They produced it. They have this protein vaccine, but it never got utilized just for variety of reasons. The upside from that adventure that they had is that they developed capacities they didn't have before.
They can do stuff that they couldn't do, and so I see very similar to this in a sense that you are acquiring facilities as you grow your product offering, to do things on a bigger and bigger scale that you couldn't do before. At the same time, you're going to probably do it profitably, as soon as you get a little bit of-- Few cost savings on the plating and so forth, this business is going to throw off cash.
Agustin Lopez Marquez: Absolutely. There's an article by ARK Invest. This is Cathie Wood's investment firm. They made a case about three years ago that whoever manages to make food delivery more affordable and healthier, that could be the size of Amazon in terms of the potential, because if you look at all food delivery combined today, it's $100 billion, Sal. If you look at the whole restaurant industry, it's $1 trillion. If you look at the whole industry including people buying groceries for eating at home, that's another $1 trillion. The opportunity really is, can that $100 billion be much bigger? As people decide to cook less at home, if they have more affordable options, that opportunity is massive.
We see a big analogy to Amazon in that sense. When you look at Amazon, the food delivery product that we're suggesting is very similar to Amazon. When you order a camera in Amazon, you don't pay 80% premium to get the camera in half an hour, you know?
Sal Daher: Oh, yes. If you go to a store, you can get it.
Agustin Lopez Marquez: You get it in one or two days. You get it at the same price as you would get it at Best Buy. For us, Sal, of all of this, we see that in the case of Amazon, to do that one day ahead delivery, they had to build the infrastructure. They had to build the warehouses, and then they had to build all the technology, and all the operations to deliver their boxes to the houses.
The way we see it here is that, actually, CloudKitchens has built the equivalent of the warehouses, you know? You don't need this massive capital investment because they have done that capital investment. Now, the rest of the puzzle, it's really hard. Robotics is not trivial. We always say it's Amazonesque in terms of the potential market size, but in terms of the path to profitability, it's not that long like Amazon where they had to make this investment and then recoup that in 10 years. At the same time, granted that the mode of Amazon, once you have those warehouses, Amazon is going to be there forever and ever.
Sal Daher: Yes. They've become part of the landscape. Agustin, we are unfortunately up against a pretty hard time. You have time constraints, I have time constraints. I'm plagued by a startup having an exit in the middle of this podcast.
Agustin Lopez Marquez: First world problems.
Sal Daher: Yes. Thank goodness for small problems. It's the kind of problem you want to have. We are going to turn this episode into an amuse-bouche. We are looking for a whole menú de degustación in the future, further conversations with you, because I want to get into why you decided to do entrepreneurship and all your adventures from coming from Venezuela to the US, studying at MIT. I want to talk about your adventures at SQZ, your adventures at nference, all these things. Those are dishes to come.
Agustin Lopez Marquez: Yes.
Sal Daher: This episode is an amuse-bouche to get us really excited. Listeners should in the meantime go to Khipi, K-H-I-P-I.com. Do your due diligence before listening to the full podcast.
Agustin, thank you very much for being on the Angel Invest Boston Podcast.
Agustin Lopez Marquez: Thank you so much, Sal, for having me.
Sal Daher: Awesome. I'm Sal Daher. Thanks for listening.
I'm glad you were able to join us. Our engineer is Raul Rosa. Our theme was composed by John McCusick. Our graphic design is by Katharine Woodman-Maynard. Our host is coached by Grace Daher.