Managing Tech in ML and AI for an Old Time Industry – Sergio Munoz, Low Gravity Ep 1.

No caption added.

By Jonathan Calmus

Feb 27, 2020

Join 15K subscribers

Listen on Spotify, Apple and Your Favorite Podcast Platforms:

On this episode of Low Gravity, Jonathan Calmus speaks with Sergio Munoz, Cosmic’s new Director of Tech. Sergio’s background is in ML and AI for document analysis and commercial audience tracking. We get into the dirty details of starting and running a new tech company in addition to speaking about Micro-Mobility and the future of the industry.

Read the transcription below!

To learn more more about Sergio check out his LinkedIn.

To learn more more about Jonathan check out his LinkedIn.

To learn more about Cosmic and to become a Mobility Partner:

To ride or rent a Cosmic vehicle go to:

Want more intergalactic content? Check out the blog

Low Gravity 1: A Cosmic Podcast

Jonathan Calmus: Hello, and thank you for tuning in to Low Gravity, a podcast on Micro-Mobility Tech and the startup hustle. Today I am joined with Cosmic Director of Technology. Sergio Muniez.

Sergio Munoz: Munoz.

Jonathan Calmus: Munoz. Sorry for my gringo accent. Munoz. Hello. Hello. Good morning.

Sergio Munoz: Hello. How are you?

Jonathan Calmus: Very well, thank you. Thank you for joining me. Its Saturday morning we’re sitting in our new offices here in Bogota, Columbia.

Sergio Munoz: Thanks for the invitation.

Jonathan Calmus: Yeah, definitely. I wanted to invite you to actually our first podcast to talk a little bit about what it takes to scale a company in Latin America, especially a tech company. And I thought that it would be an interesting conversation with you because you, yourself are a founder, you yourself have started a company that was right at the verge of exit as well. You guys raised some money, you got some clients. So tell us your story a little bit.

Sergio Munoz: Okay. I had a company that we are doing products with mature learning. We start a project doing facial recognition to give places that have an amount—a lot of amount of people around, but they have to recognize only a certain amount of people, like people of interest. Like if someone has like a black list or white list for VIP kind of stuff.
So at the beginning it’s like we get an interesting product. We go with clients, I’ll say, okay, let’s test this for free. We start trials that month, and they said, yes, your product is really nice. But then they didn’t decide to buy the technology, not with us neither with the competition that also they were trying to sell like big company, international companies.
So at start up, you start to bleed if you like spend money, spend money, like try to move to the client to buy. But also if you miss your traction, you will say, okay, you were testing your product with five different clients. They all say they like it. So you say, okay, I’m incorrect, but I’m having traction. But at the end if they don’t buy it’s difficult.
And I think that happened, because in security market is like kind of, they move slow.

Jonathan Calmus: This was in Chile at the time?

Sergio Munoz: This was in Chile. So then we start to pivot So we changed to working, recognize and work with documents, classified documents named entity recognition. Just to do some stuff for automate back end operations and in their first demos we get pay. So, Oh, nice.

Jonathan Calmus: So now your business model actually has potential?

Sergio Munoz: Yeah. But also we were like running out of cash, we need like Oregon investment. I get an offer of three groups and I decided one to work with, because they showed that they have the clients. It’s like we always recognized our self that we were good at putting the tech but we recognized some lacking in commercial skills.
So I say okay, this guy have figured out their market. They have difficulty to get their product on astroid so it should be a good fit. So I think the first thing that is really difficult, and you can believe today if you’re really don’t gets pass the—like know that your product fits the market.
And you can read about in books on everything. But really it’s like, if you can like test and really know the potential of pain without beginning to code before I think that is the best. I don’t think that you can always do that, but if you can try and achieve that, I think that is one of the first steps to take.

Jonathan Calmus: Reduces your risk a lot. So that’s an important point. Because a lot of our—-hopefully a lot of our listeners. Currently it’s the first episode so we have no listeners. But hopefully a lot of our listeners are considering starting their own Micro-Mobility Fleets. And Micro-Mobility I think is a great example of what you’re talking about because it’s such a capital intensive business.
You don’t just need the software side. You don’t just need to find the proper hardware that could actually live in the circumstances of being on the side of the street. But you also need to go and do the commercial parts and do the public relations parts and make sure the government is giving you proper access to the streets and everything like that.
So it’s a very important point and this kind of comes to the crossroads of most early stage entrepreneurs build or buy, right. From the first day all of us have this whim that okay, the exit comes from the proprietary code and at the end of the day that’s not always true. I mean there’s a lot of companies that prove that.
I used to be in the Bed In A Box Space. Basically we would compress mattresses, these very high end memory foam mattresses. And you’d ship it in a box cross country in the United States. And the company tthat was really pushing and kind of got the biggest was this company called Casper. And they’re on the verge of IPO right now. But their story is super interesting because Casper doesn’t own a single machine, they don’t own any factories. They cut deals that they were able to produce their inventories prior to ever holding any inventory. So everything was on demand. Most of it was drop shipping in the beginning. And they’ve scaled to a huge level Silicon Valley level of scaling without ever owning any proprietary asset. Maybe now it’s changed, but at least at the time. And starting with very low inventory, which is kind of contrary to what the impulse of being an entrepreneur is, we always want to build stuff.
I think that’s what’s fun about being an entrepreneur to a certain degree. It’s the new word for inventor. You know, like 50 a hundred years ago we used to call it inventor. And now it’s entrepreneur. It’s a different, a little bit of a different word. Well that’s really interesting.
So you have a really strong background in ML and artificial intelligence. Let’s talk a little bit about that. Where do you think Micro-Mobility is moving? And how do you think it’s going to be leveraging artificial intelligence in ML? What are your predictions on the future?

Sergio Munoz: I think there are maybe two topics that I think artificial intelligence will really help. One is regarding geographic information. Because in Micro-Mobility you want to put the asset near where is the demand and those are really complex model. You have companies like Uber that this kind of stuff is kind of core of the business because that is why you thrive, also the venue. If you can optimize their supply near to the demand. It’s like you already discussed. So that is broadly one thing that is kind of more mature that the current mature learning algorithms can come deal with that.
And the other part is more like put intelligence on the device, meaning step two, if the demand is far from the device, can you tell the device go there without having a human to move over there?

So that I think like today’s a little bit more science fiction. But I think that big breakthrough will be a game changer. And you have a lot of big companies working on autonomous driving. So that is the future.

Jonathan Calmus: So not only using and leveraging AIML to know where the demand is, but also even potentially autonomously going towards the demand and being able to redistribute and rebalance your fleet, without having too many humans involved in the logistics side of it?

Sergio Munoz: Exactly. And also it will be a game changer in last mile delivery.

Jonathan Calmus: Yeah, definitely. There’s a few companies —–there’s actually a Colombian company we were talking about the other day. They’re called Kiwi. A group of really smart guys. Right now they’re in Berkeley I want to say somewhere in North Cal. And they’re working on autonomous last mile delivery for the purpose of like a RAPPI or an Uber Eats or Postmates, but using a little autonomous robot and it’s very nice. But I think one thing that is very important is even with the big companies who are working on these types of tech they really require very specific environments. Like the streets of LA couldn’t really handle an autonomous scooter everywhere, simply because the roads are kind of messed up.

Sergio Munoz: Amazon is working with drones.

Jonathan Calmus: Yeah. On the drone side. But I’m saying on the ground, like you’re not going to have a drone pick up a scooter and move it to another place.

Sergio Munoz: Why not?

Jonathan Calmus: For obvious reasons. I hope for obvious reasons, I hope. But if this thing is going straight somewhere and it relies on a flat surface it’s going to end up toppling over. So you see most of the implementations which is still fascinating that finally, science fiction has caught up. We see little robots delivering food and stuff. You really primarily see them on a university campuses more than anything. One because the demand is so high, a stoned student wants their burrito at two in the morning, but also because it’s very predictable environments, you can actually guide yourself there.

Sergio Munoz: But probably in the future this device can use like Elon’s Tunnels in LA.

Jonathan Calmus: The Boring Company probably, and I’m sure there’s some grandiose scheme. But seeing the trend of where everything’s going is very interesting. I think it’s the most both promising and challenging part of Micro-Mobility is the optimization of logistics in general. I mean, we know as Cosmic and part of the reason why Cosmic took our approach to becoming a platform and a kind of franchise model is because we really wanted to focus on optimizing those areas so that an operator or an owner could easily go to market. Not really think about those other details. But that’s where all the costs goes. It’s between a–you basically share the cost between the loss of the value of the vehicle and combined with the cost of humans running logistics.

Sergio Munoz: But seeing the future of Micro-Mobility and Shared Mobility is something that Latin America has kind of similar. New York, Los Angeles, you have a huge amount of people in a really narrow space and traffic is awful. It’s like myself, years ago I always drive a car, but it gets to a point when traffic was so horrible and you have to pay parking and expensive parking everywhere. It’s like you say. I don’t want to drive it anymore but also you don’t have too much options. Like now we are getting as you say, a city has to become more friendly. Like in the street you have a scooter or bicycle, a street only kind of stuff. But also like owning—but sometimes you need a car. So if it’s more easy that, okay.
And like in the weekdays I move with my bicycle or I rent a scooter. But on the weekend I need a car, about 3 hours. And I don’t want to go to the regular rent a car that they asked me—like the process is slow. I have to go to pick the car somewhere far from where I am and then deliver it back. So there’s a lot of opportunity to bring it. It’s about experience. To narrow the experience of sharing.

Jonathan Calmus: Options are changing the landscape in general right now. I think one thing I was observing, I kind of had this like mini fake epiphany the other day where none of this would’ve been possible in another time. And it’s probably an obvious thought, but consider how many technologies had to mature at the exact same time or at least on the same trajectory in order for us to be doing this.
Like the other day I was talking to our data SIM provider. Consider how incredible of an achievement it is that one single SIM card can be able to connect to a network worldwide. Consider how incredible of an achievement that we took this huge cellular technology, shrunk it down small enough and now we can actually attach it to bicycles and scooters and mopeds and cars and a variety of other vehicles and devices in the world.
Everything is moving at a very similar path in a very similar velocity as well. And so everything is coming to a head at the same time for us to be able to take advantage of this. And probably that’s also tied to the first thought, why the kind of conversation of build versus buy because it’s an exciting time. I mean, every single day you see someone else created a new technology built on the backs of giants, right? Like you take these crazy complex ideas and concepts and you shrink it down because now there’s access to things like APIs and STKs and micro services of other people. And I mean, there’s all these different opportunities and it probably becomes a little bit easier to forget how difficult it is to actually create something. In some ways it’s easier and in some ways it’s even harder now, to create.

Sergio Munoz: Actually it’s like as you say, having worldwide SIM card it’s a really huge advance. Still we need a little bit more mature of IOT, more a part of the vendor because protocols are getting better.

Jonathan Calmus: Who do you trust to buy this?

Sergio Munoz: But if you’re looking IOT from China, they are still relying on older protocols. So there’s a little gap there to narrow. Because one is that it’s difficult when you have to manage the device, you want them to be connected. There is a big pain where they get disconnection. There is a kind of a reality to the today in IOT environment.
Probably 5g and Google help in some way. When it’s well implemented in dense cities because also like a currently you have blind spots, regarding buildings and that kind of stuff. So I think we are going to the right pathway. The technology is kind of a exciting moment that we are in avert of a really big change regarding like technology be in—that with the change of having technology we are able to do more incredible stuff.

Jonathan Calmus: I think because Micro-Mobility became so prolific, so fast and you see them everywhere. You see shared bicycles, you see shared scooters, kiosk based kiosk lists. Really the way that most people see the industry is really only three years old. I mean it started with the boom of Bird and Lime in the United States. Not even taking into account, you know, all of Asia, China specifically with Shared mobility. But it becomes maybe very easy to consider this some very mature space. And the reality is it’s just not, it’s the first days.

Sergio Munoz: What do you sing about Lime? Like going out of markets is like the bubble is exploding or is getting more mature?

Jonathan Calmus: Just understand what the different paths are to this industry. The leaders in the industry took a centralized approach in the beginning. What do I mean by that? Bird and Lime were purchasing their vehicles directly and launching in multiple Cities at the same time. And they kind of broke the rules of how most of us were taught to build a business, go and saturate one place at a time, pick a country, do you know the best you possibly can get the highest user ship you possibly can. And then only at that point jumped to another place. And this took much more of the block chain crypto growth path in terms of the hype and the excitement around it. And so I’ll answer through answering through Blockchain. Is Blockchain dead? No, not at all. But the big over explosion of excitement. Yeah. That part of it kind of dwindled down a little bit and for obvious reasons.
So when these huge mega corporations like Lime or Bird enter or leave a market, it doesn’t actually represent what’s happening locally. It really represents what the corporate strategies are. For Bird or Lime you’re talking about companies sitting in California. So when we go and analyze a market in Mexico, for example, which we’re, just entering now or if we go and analyze a market in Chile or Africa or some of these other emerging markets.
You can use Lime leaving Santiago and Buenos Aires as an indicator as there’s something wrong or the company just grew way too fast and they have to balance their numbers because it seems like they’re also on the verge of an IPO, both of these companies.
So there’s different games being played. It’s really hard to pay attention only to the little media signal and say, okay, this is what it is. And because Lime‘s leaving its the end of an industry. No, not at all. This industry is growing really fast. There’s a lot of markets that have not been saturated yet. We do our best in these kind of underserved, underprivileged, but also high tourist destinations. For example in Pucon I mean, our ridership is incredible right now because it’s summer season and Chile and a lot of people are traveling and things like that. But normally you’d never find a Lime in Pucon it’s just a weird environment for that. Everything’s really just based on strategy and the mega corporations went after the metropolitan city.

Sergio Munoz: You hear about the term Blitz Scaling?

Jonathan Calmus: Blitz Scaling? No. What’s that?

Sergio Munoz: It’s a book. It’s like the book that subtle, the way of really fast grow with like kind of Airbnb what they did. Lime is like—once you get your prototype and you get your market fit then you get a lot of money and you like cheap to—saturate.

Jonathan Calmus: Everywhere–saturate.

Sergio Munoz: t’s kind of a really risky play because you are like spending huge amounts of money. It’s like this, you’re buying the scooter to put hundreds in several cities the same time. And then in the tech part you have to pray that you can catch up.

Jonathan Calmus: Things are working and that they’ll get better. Yeah, no, I agree. It really just depends on strategy. Most of the Cosmic partners are on the enterprise side looking to change the landscape of mobility. And change the landscape of how their company interacts and deals with the problem of environment change. So we have a lot of sustainability partners, a lot of people who are trying to implement this through their corporation.
And their intentions are less about money and less about revenue. And their intentions are more about kind of having a change in perspective about their company. So we get a lot of utility companies working with us. We get a lot of gas companies. On the other side of that our partners are looking for profit. And if you’re comparing the profit—-the person with the intention of profit versus the Lime or the Bird, it’s just a totally different strategy.
Because the way that Bird and Lime actually achieved true profit, meaning being the cash cow business that we’re all looking for is that they go into some form of an exit. You know, fundraising can only take you so far, even if you’re raising billion at a time. Their intentions are just very different. So that’s why their strategy is that.
But the operator that we have and the partner that we have in Pucon, their intention is very clear. They actually want this to be a real business and they want to generate money from it. And I think it just really depends on what side of the spectrum you’re on. And I’ll use the example of Casper as well again, because I think it’s kind of a fascinating story that a mattress company emerged out of nowhere. I mean number one it’s not a sexy business. And number two mattresses, especially compressed mattresses is very old. It’s at least 30, 40 years old the compression of mattresses and somehow they broke the mold and somehow they said, no, this is going to be the next unicorn in Silicon Valley, is this mattress company.
But they don’t make money. Really. I mean that’s just the reality to get to that scale in such a short amount of time. You’re taking big projections on the future and you’re not focused on the day to day recuperation and income that you have to be as a smaller local business. Everything is really just based on direction and strategy and everything else. If you plan to be a huge bird, you have to raise $100 million. There’s no way around it. If you want to be an operator in a city servicing your community and getting people excited and helping move people around and generate profit around that, then there’s other ways to attack the business. It’s the difference between trying to start McDonald’s or buy a McDonald’s.

Sergio Munoz: I get a question for you, about your opinion on businesses in Latin America. It was like you say, if you’re in the U S and you want to grow these startups. These guys get million and billions. What do you think? You have raised money in the U S and also in Latin America. Do you think that it’s possible to be in Latin America and try to achieve that kind of investment to grow as fast as if you were in the Silicon Valley?

Jonathan Calmus: Yes, but right. It’s always that type of answer. Yes. But because you really need to have a strong network and you really need to have a strong track record. So for example, Grin. It started in Mexico and they grew throughout all of Latin America. They were very good at doing mergers quickly. So they ended up becoming at a similar scale of Lime or Bird exclusively in Latin America. Raising big money in the typical Silicon Valley Tel Aviv style path of raising money.
But they’re also an anomaly in the sense that they have access to all of those networks. They have a very good track record. The guys who started it come from the VC worlds themselves and it’s just a different story. So I think if you’re sitting in Latin America and you don’t know the full context of the story, you kind of trick yourself into believing, yeah, that’s how it is.
One thing that’s really different in my experience, I mean, I’ve only really been doing business in Latin America for about four years now. But the biggest difference I think between Western Culture and Latin Culture is trust and confidence. In the U S deals tend to move relatively quickly, contracts even though they’re a blocker, they’re never a blocker enough for some form of a deposit or let’s get started or something like that. Whereas in Latin America, things are much more bureaucratic. Things are much more structured. And if you don’t have those basic things, even just at the employee level. In my previous companies, I was able to start and have contractors and employees working for months or sometimes even a couple of years just on a handshake because of the confidence, that we’re all there for the same reason.
In Latin America, I’ve experienced people being very rigid and saying, no, I will not even start until I get my contract signed by both parties. And again, if your able to tap into the right networks of people where you have track record and you’ve demonstrated it, it ends up moving very quickly. If you’re unable to do that, you get a really slow and very difficult and painful process. I was used to raising money in the States and raising money in the States. While it’s not easy, was easier, at least in the beginning for me, primarily because I didn’t have a network here, but once my network started building here, I started realizing a pattern of certain types of characteristics from VCs and investors here. You will never, ever, ever get a no ever, which is crazy. You’ll get a lot of smiles and then you trick yourself into believing, Oh, this is great feedback, they’re smiling.
Because every investor meeting I’ve ever been in, you always deal with someone who has a little bit of a poker face, right? Even if they love what you’re doing, they want to hold back their excitement just to make sure that they’re being, as you call a sophisticated investor, thinking about it before really committing to it. In Latin America, you do get feedback. People are smiling at you and they go, I love this. This is a genius idea. You’re going to be the next Steve Jobs. You’re going to be the next… And they just kind of tell a story.
But that’s also the culture here too. It’ very much a storytelling culture. It’s very much a friendly culture. I think that’s what attracted me to Latin America. So on the business side, that becomes really difficult because if you’re not conscious and sensitive to the differences of the culture, you’re in a position of tricking yourself and you wake up every morning going, today’s the day we’re going to raise our seed round and it just never comes around.
It’s almost kind of funny. As a westerner, I would recommend raising money in Latin America and as a Latin, I would recommend raising money in the United States. Because the grass is always greener and everyone wants something that’s from a way. Because they know what’s here, you know? Oh, it’s the same stuff every day we get pitched the same stuff.
But there’s something novelty to an accent. My dad taught me this. My dad’s from England and he lived in Los Angeles since he was like 15 years old. And he said that just the moment he came to United States, his IQ increased by 15 points. Just because of his accent. And it’s totally true, in Latin America, even though there’s a joke around being a gringo, its novelty. In United States as a Latin, if you’re doing tech, it’s a novelty because it breaks the mold. I mean, we all know the biases of the rest of the world of Latin America and the kind of stereotypes.

Sergio Munoz: You have something stories, like for example, corner shop, their first money came from the U S.

Jonathan Calmus: U S exactly.

Sergio Munoz: Not from Latin America.

Jonathan Calmus: Cornershop also are really unique. Cornershop is not Latin, Latin. It’s a group of outsiders, but still its…

Sergio Munoz: And also they had a track record before off of exit. So yeah.

Jonathan Calmus: And that’s what it’s all—I think that’s what we’re all working towards. It gets easier to pick up the phone.

Sergio Munoz: But look at that example, they already have the track record and for Cornershop, they also get the know in Latin America.

Jonathan Calmus: Yeah.

Sergio Munoz: And they had their right to their track record.

Jonathan Calmus: Yeah. I can’t remember who told me this, but startup land is just about patience and perseverance. The faster you try to move in your head, doesn’t mean you don’t have high output and doesn’t mean you’re not fast in terms of actual delivery. But the faster you try to move in your head, the slower you end up becoming in especially in Latin America. Because you keep skipping steps and bureaucracy means that the order of operations is very important. You have to do this set of tasks before you can go do those sets of tasks.
So for example, when we raised money, the typical question was, have you formed your company? Have you done this? Have you done that? And to me that was a kind of weird thing because I’m like, yeah, that’s the first thing that we did. Even though I wouldn’t recommend all startups to go and form a company overnight, you don’t have to spend that money for no reason.
But because we were already kind of accelerated, it was kind of second nature. Yeah. Obviously I would set up a corporate structure. Obviously I would create the legal infrastructure around all of this. And I guess they were just dealing with a lot of people who have never done that. And everything is about context, especially in front of the people you’re in front of. And this is what most sales men in Silicon Valley and self-help talks about, knowing your audience. It’s really important to know your audience. And as a foreigner and as a Westerner coming into Latin America. I think the biggest learning curve I had was it was hard to read my audience for the same reason. I’m saying people are sitting there smiling. To me that’s a good indicator. And in reality that smile translates into them never answering the phone again. It’s a different experience all together. And then it got double crazy. Because Micro-Mobility got so saturated so fast that they started probably getting a lot of pitches, over and over and over again.
Hey, there’s this company Grin and there’s these gringo from United States coming in and Uber’s here and now DD from China’s in Latin America. And all of a sudden you have a lot of indicators that globalization is not just happening from Latin America to the United States, but it’s worldwide to Latin America.

Sergio Munoz: Yeah, that’s true.

Jonathan Calmus: And it changes the landscape, but I don’t think the VCs caught on as fast as they should have in LATAM. I think, I don’t know. Maybe they just didn’t change their philosophies and frameworks. Because the same speed of creation is the speed of funding. There’s a reason why most people flock to Silicon Valley or to Israel or to environments that they know what the funding thesis and philosophies are of investors. But in Latin America, if you take six months to make a decision, you’ve actually not only lost your opportunity, but you’ve potentially lost the opportunity of the entrepreneurs themselves.
It’s a weird concept because it’s really easy to feel desperate and lower value when you’re pitching to a VC and you’ve never done it before. Because in your head you go, shoot, I really need money. If I’m going to make this company, this vision grow and succeed, I’m going to have to fund it somehow. And at this point, I don’t really care who I’m going to take. And it’s the biggest mistake you could possibly make when you’re raising money.
Because as much as the VC is vetting you and deciding if you’re good enough for them, you have to be doing the same for the VC. There’s options out there. It’s never just one person or one organization. But because they’re the one writing the check, we kind of all trick ourselves. Remember, it’s not them writing the check, it’s us selling our company.
It’s a very, very different logic and philosophy there. I don’t know, I’m not really sure about that. Awesome. Well said here. Thank you so much for joining me today. We could talk forever about these topics, I’m sure because they’re very fun topics and, everything is kind of interconnected. But I really appreciate your time today

Sergio Munoz: No. Really thankful for Invite. I’m really glad to also share my experience with everyone.

Jonathan Calmus: Yeah, it’s really good. If you have any comments, questions, if you want to ask anything of Sergio in the tech side or more the business side and startup side, feel free to message us in the comment section below. And if you’re interested in launching your own fleet, go check us out at