The Stack Overflow Podcast

We bought a university: how one coding school doubled down on brick and mortar

Episode Summary

Paulo and Guilherme Silveira, brothers and cofounders of edtech platform Alura, join the home team for a conversation about polyglot programming, edtech, and the role of generative AI.

Episode Notes

Alura is a Portuguese-language edtech platform where users can learn programming, backend and mobile development, data science, design and UX, DevOps, and more.

They started small, grew into a bustling online program, then purchased a majority stake in FIAP, a private university in São Paulo, Brazil.  

Paulo and Stack Overflow Director of Engineering Roberta Arcoverde cohost a popular Portuguese-language podcast about programming, design, startups, and technology.

Paulo’s new open-source project is full of career resources for T-shaped developers.

Connect with Alura CEO Paulo Silveira on LinkedIn.

Connect with Alura Chief Education Officer Guilherme Silveira on LinkedIn.

Connect with Roberta Arcoverde on LinkedIn.

Today’s Lifeboat badge winner is netblognet for their answer to Get JSON object from URL.

Episode Transcription

[intro music plays]

Ben Popper Hello, everybody. Welcome back to the Stack Overflow Podcast, a place to talk all things software and technology. I'm your host, Ben Popper, joined as I often am by my colleague and collaborator, Ryan Donovan. How's it going, Ryan? 

Ryan Donovan Hey, Ben. How are you doing today? 

BP You know, I'm getting by. We have some great guests today: Paulo and Guilherme. They're joining us through a friend of ours, Roberta Arcoverde, who's been on the show many times, is a very senior engineer here at Stack Overflow, and has been responsible for building some great stuff which we have had up on the blog recently. So I think today we're going to touch a little bit on EdTech, we're going to talk about what it means to be a polyglot programmer learning in different languages, and we're going to talk a bit about the world of startup versus industry in academia, so I'm very excited for it. Paulo and Guilherme, welcome to the show. 

Paulo Silveira It's a pleasure to be here. I’m really honored.

Guilherme Silveira Thank you for inviting us. 

BP So for folks in the audience who don't know, tell them quickly, Paulo, you're on a podcast with Roberta that's in Portuguese and is, in Brazil, big among the software development community? 

PS Exactly. It's quite big here in Brazil. We have been doing this for the last six years, and Roberta joined us five years ago, so it's a pretty [big] hit here. And maybe you don't know, but Roberta, people will get to her here in Brazil to take pictures with her.

BP She didn't let on that she was that big of a celebrity. I'm going to have to tease her about it now. 

PS She is. 

BP Yeah, next time we're at the company meetup I'll have to get a picture. 

PS [laughter] You should. 

BP And so, what's a little bit, for both of you, sort of your background? How did you get into software, and what led you to doing this show together? 

PS So me and my brother here, we co-founded a coding school, just like so many others that we have nowadays, that we usually will [call] EdTech. And this has been running for a while, [but] for the first time in a traditional brick and mortar style, and for the last 10 years online which is called Alura. And we are pretty big here in Brazil, and we started just like Joel with Stack Overflow. We started inside the community, inside the web forums, much before any kind of Facebook groups, WhatsApp, or even Stack Overflow. Back then, at that time everything was decentralized and we started there from inside the community. 

RD Yeah, one of those classic PHP forums, right? 

PS Exactly. Web forums, PHP forums. That's it, Ryan. 

BP And so tell us a little bit more about the company that you two created together. What do you focus on now and how has it developed over time?

PS So in the beginning we started with the cool stuff back then. Agile and Java were not mainstream, so 2000, 2004, 2010. These kinds of technologies were for people who do not want to work, hippies and hipsters and so on. And that's how we started, that's how we got big, because no one was looking at this kind of thing. They were betting against it. I worked for Sun Microsystems and every kind of course was Java related. There were no Apache or any kind of open source related courses because the big companies were fighting against it. And 10 years after that, we made the right bet. It was not a bet, just to be honest. We just wanted a cool job and we wanted to be professors and PhDs so we were studying and having fun with the coolest technology back then. 

RD What has it been like making the jump from having the brick and mortar school to having a big online presence? 

GS So it was quite a fight because we started brick and mortar with just one small room and six computers. So that's how we started and we took a lot of care about how we presented the teaching methods, the content and everything else. So that was growing quite slowly at the beginning– 2004, 2005, and so on. And we always had, not fights, but we always argued about if we should be teaching online or not, because we had the numbers and we knew that people all over Brazil, and some people from outside Brazil, wanted to learn with us, and they would just come during their vacation or something like that. They would come to São Paulo to learn with us. So we opened in Jundiaí and Brasília, in two other cities, but it was still not enough. At the same time, it's quite a headache to have several centers and to take care of quality and everything else, so for quite a while it was just an argument back and forth. Should we do it? Should we not do it? Do it, don't do it. Always a question regarding quality. And at that time, at the moment that we started doing the online, it was because there was one platform Livemocha, which died, it was bought by Rosetta Stone later and died, and it introduced quite a few social features to online learning, and that was what showed us, “Okay, we can do something different besides just a video that you hit ‘play’ and you watch just the traditional thing and that's it.” So that was the moment. 

BP You have your learning community online as well, and that kind of keeps you going or allows you to learn more. I mean, that's kind of one of the theses of Stack Overflow– learn from a trusted community and interact with them. You're going to get a lot more out of the wisdom of the crowd than you will on your own. 

PS Yeah. And 2011, 2013, that's when we started doing the online business here. You should imagine that the situation is pretty different from the US. So in Brazil there was almost no broadband connections, no smartphones besides São Paulo city center. So it was even a challenge to have this ramp up back then, so it was a bet. And we had the innovator dilemmas because we were threatening our own business with a product with smaller margins, and if people loved it, it would kill the first business. So we had exactly this situation, which when you are inside it, it's hard to make a move. It's really hard to say, “Hey, let's go this way,” so we started adding some courses and content that we did not have as our main business, and maybe we should have moved faster back then.

RD I mean, like you said, that's got to be hard to kill your own business like that. But obviously there was an appetite if people were coming on vacation to take courses before that. 

[music plays]

BP All right, everybody. Today's episode has a very special sponsor– yours truly, Stack Overflow. Now we all know the frustration of searching for answers on internal wikis that have gone stale, or trying to find that one email or chat thread from months ago with the information you need to get unblocked now. Well, there is a better way. Stack Overflow for Teams is a knowledge base that has all the features you already know from stackoverflow.com, but reimagined for your organization so that you and your teammates can collaborate, quickly find solutions, and be more productive. It's like a private Stack Overflow for your organization's internal knowledge and documentation and it's used by companies of all sizes like Microsoft, Expensify, Bloomberg, Dropbox, and many more. You can always try it out for free at s.tk/teamspod. Go over there; let them know the podcast sent you, s.tk/teamspod. If 50 teammates won’t cut it or you need more advanced knowledge management capabilities, head on over to s.tk/teamsplan and use the promo code ‘teamswin’. You’ll get a 30% discount on the basic or business plan for your first year, courtesy of the Stack Overflow Podcast. Alright, spiel over, let's get on with the show.

[music plays]

BP And so at what level now do you teach courses? I know one of the things you wanted to talk about was sort of learning on your own versus learning in academia, what does it take to get in an industry? Do you have folks who are learning from you and going straight into a job and skipping their computer science degree? 

PS Yeah, I think this is a great discussion, Ben, because 10 years ago when all these big EdTechs in coding started, I would say Udacity, Udemy, CodeSchool, Pluralsight, and so on, there was this kind of hype that you don't need university, you don't need a diploma. You can skip it and get a good job in tech. Which partially is true, but a few years later, nowadays we can see even in these big tech FANGs, we'll say that you don't need a diploma to get a job here. If you get there and see, most of the developers, I would say more than 90% have undergrad degrees and even masters and PhDs at top-notch universities. So this kind of battle is not a battle here. These are different things that are for different moments and maybe even for different kinds of jobs. So I can bet that you both have– well, just to be honest, I took a look at your LinkedIns. So here we all have university degrees and I can bet you’ll say that I learned a lot there, some stuff that I don't use day to day in my job, but most of the things that I learned then, the social connections, how to think, how to criticize thinking, I use even today. We also made a big move last year. We bought the most famous private university in tech here in Brazil, because we always saw that that was our first plan– to become a university one day, because we see that there is a path here. When you start a job in tech just doing usual coding, I would say that just doing apps and basic stuff –it's quite hard to say basic stuff– is one thing, and doing computer skills and computer science is another, and there is space for both and maybe one day even the ones that started without a diploma will need to get deep tech knowledge. 

BP I have to say, I'm not sure why, but I don't think I've ever heard that before where a technology company went and bought a private university. Certainly I know that there's big tech companies that have basically built their own universities in essence to onboard people to their platforms and technologies and give them cloud certificates, but that's so interesting. So you bought a physical campus and a bunch of tenured academics and all that? That sounds like the opposite of a startup in a lot of ways. What was that experience like? 

PS It's a wild ride right now because the investors and even the market was saying, “Hey, you are going the other way.” And we think it's the right way. If you take a look at the numbers in the US over the last 10 years, the undergrad course that has grown the most these years is computer science and computer engineering. And then it comes nursery and psychology and all this kind of stuff that I think are really practical, and then things that you can have a high impact in a short time [with] during your job. So for me, it makes perfect sense to have a physical university where things can happen and where you can exchange ideas and social skills and practice what is more human-like. That's exactly our bet. And I think [with] all the social features that we have in Stack Overflow and other places [that] this is the key differentiator. 

RD That's interesting. I've talked to some programmers who run this group, Papers We Love, where they talk about academic computer science papers and they all originally started with humanities degrees like History and English. And I think it's interesting getting back to the college campus to get that cross-pollination, to get that thinking about how to think idea. Have you thought about any courses that are working with that cross-pollination? 

PS Well, anything that we can get to algorithms, data structures, and data science, and you can spark some kind of interest in your students. Because even if you are doing Excel and Google spreadsheets and people are taking a look at big numbers and making a decision, you can get them to say, “Hey, if you do this kind of search, or if you don't do this kind of analysis, you are losing a lot of time. This is not a good practice.” And so I really think that there are a lot of different ways to show people that the underlying structures, algorithms, and even operational systems that nowadays people usually don't look at, there is a lot of stuff that people can learn from that. There is a friend of mine that says that to be a senior engineer or developer, you need not only to understand your stack, Java or JavaScript or C#, you need to understand one or two layers downstream. So you need to understand the virtual machine and so on, and even the operational system or the browser. So to get to know this kind of thing deeply, usually the university style, the technical papers, computer science conference is the best way. It's not the only way for sure. There are a lot of developers that have never been inside a university and they are 100 times better than us, but these, to be honest, are exceptions. 

BP Right. I think one thing you said there that we see a lot on Stack Overflow now is that the power of statistics and algorithms and data science is increasingly being used all across academia and so we have all these stack exchanges for other sciences and disciplines and they all want to learn R and Python and figure out how to do that stuff for themselves because it's useful for their research for their academic career. So it's definitely something that, like you said, gives you additional tools as an academic. I have to address the elephant in the room. We're talking about going to a brick and mortar college or going online to learn coding skills. You mentioned, and I understand why this is true, that the number of students going for CS degrees has been growing because it seems like a field where there was never enough skilled people to meet demand, and you could come out of school and land a high paying job. In your opinion, what does that world look like in a year or two when anybody can turn to an AI assistant and say, “I'd like you to write some code for me.” It does. You run it and it has a little problem and you say, “Debug this for me,” and it does. And you always have this private tutor or assistant that can act as sort of like a second brain. What does that mean for all the kids who are still two or three years out from graduating when they get out and want to be that entry level engineer? Do you think the world will have changed for them?

GS So definitely it's going to change and the question is how much. Today, what we can do with those generative AIs is it helps us whenever we know the path that we are following. So if I need to do something, I have to be as clear as possible. And we have worked with clients in our lives and we know that nobody knows what they want, not our clients and not ourselves, or we don't know exactly. When we start writing our code, we don't know exactly what we want. And when we write it in Portuguese or in English or whatever, we even add this new layer of failure that we are not exactly transcribing what we think we want, and the computer will execute what it perceives what we think we want, what we think that the client wants and so on. So it's exactly as you mentioned, Ben. It's a tool. So we need to learn how to use the tool. So is it going to help us to write a hundred percent of our codes? It doesn't seem like that in the near future. It doesn't look like that. Does it help us writing code and solving smaller problems step by step, guided by our knowledge? Definitely. It definitely looks like it can be used like that and that we should be learning how to use and make the best use of that in that way. 

PS Yeah, I remember when Sussman and MIT changed from Scheme to Python, the universities were all talking about, “Hey, you are going to this kind of mundane thing that is Python. We should stick to Lisp and functional programming.” And there is a quote where he says that, “Hey, look, the computer thing is changing. 20 years ago, we needed to write from the ground all the algorithms, all the data structures, all the math things,”and nowadays it's more like a puzzle where you are connecting pipes and using the right tools.” And even the modern stacks where you have all these kinds of things for C#, for Java, and so on, it’s again, the same thing so you don't need to care about some stuff, although when you get to really big problems, you need to understand the core of the code of what you are doing. 

BP Yeah, exactly. So my experience has been that it's exciting for me as the world's worst programmer to get to use it. I use these natural language prompts. It's giving me things to do. I can get pretty far, but then there's always something, Guilherme, like you're saying, some friction where it's not that the computer is giving the wrong instructions, it's that I’m executing them incorrectly outside of the chatbot or the IDE. And then I end up at Google and then I end up back at Stack Overflow. “Oh, you’ve got to zip the files like this.” Oh right, okay. Got to zip the files. Just those little things that, like you mentioned, you would probably learn in a human interaction at the office or at school or whatever it may be. 

RD Yeah, it’s trained on solved problems so it can only really get code for solved problems. One of my writers is working on a piece that the programming isn't the hard part, it's the requirements, and I think that a lot of people are thinking about that. This code is coming out, it's working. Does it solve new problems? Is it what we want?

[music plays]

BP All right, everybody. Thank you so much for listening. As always this time of the show, we want to shout out someone who came on Stack Overflow and saved some knowledge from the dustbin of history. Awarded March 17th to netblognet, a Lifeboat Badge for answering the question, “How do you get a JSON object from a URL?” Well, netblognet has the answer and has helped over 585,000 people over the years. We really appreciate it. I am Ben Popper. I am the Director of Content here at Stack Overflow. You can always find me on Twitter @BenPopper. If you have questions or suggestions for us, send us an email, podcast@stackoverflow.com. And if you like the show, why don't you leave us a rating and a review. It really helps. 

RD I'm Ryan Donovan. I edit the blog here at Stack Overflow. You can find me on Twitter @RThorDonovan. You can find the blog at stackoverflow.blog. 

PS And I’m Paulo Silveira here from Alura, CEO at Alura. I don't have many links and content in English for now, but you can take a look at techguide.sh. It's a new open source project that we are running to map careers and content about what you should or could learn as a T-shaped developer. It's a new project that we are running and we have it for some careers in English. 

BP Oh, very cool. We'll put the link in the show notes. Go ahead, Guilherme.

GS And I'm Guilherme Silveira. I'm Chief Innovation Officer so I do research and content for Alura and whatever else comes into my lap, and thank you so much for having me here. 

BP All right, everybody. Thanks for listening and we will talk to you soon.

[outro music plays]