Serverless is your competitive advantage

Nader Dabit is a web and mobile developer, who specializes in building cross-platform and cloud-enabled applications.

We also talk about:
  • how he managed to build a following on almost every popular social platform,
  • how he got started with his own training company focusing on React Native
  • what serverless means, and why you should care about it,
  • how to build an MVP using a serverless-first mindset,
  • and how frontend developers can leverage serverless technologies to become a full-stack developer.
Picture of Nader Dabit
About Nader Dabit
Nader Dabit is a web and mobile developer, who specializes in building cross-platform and cloud-enabled applications. Right now, he works at Amazon Web Services, where he develops features in the client team and improves developer experience. Before, he founded his own training company, specializing in React Native, and trained engineers from organizations such as Microsoft, Amazon, the US Army, and many more.
Today’s episode is sponsored by CodeSubmit – the best take-home assignments for your tech hiring!


Read the whole episode "Serverless is your competitive advantage" (Transcript)

[This transcripts is the result of a community effort. You can help make it better, and improve the podcast’s accessibility via Github. I’m happy to lend a hand to help you get started with pull requests, and open source work.
Special thanks to Jack for helping improve this transcript.]

Michaela: [00:00:00] Hello, and welcome to the Software Engineering Unlocked podcast. I'm your host, Dr. McKayla, and today I have the pleasure to talk to Nader Dabit, a senior developer advocate at AWS. But before I start, let me tell you something about coincidence. Sometimes the right things come at the right time, I believe. Three weeks ago, the founders of CodeSubmit, a take-home assignment platform, reached out to me to sponsor the podcast. They showed me the hiring tool and I was really impressed. So, we made the deal. But do you know what happened at the same time? I started to work with a startup where I need to hire a team of engineers. So, this week I used CodeSubmit to create a custom-made system design task. The best part - from my perspective - of the tool is that it natively integrates with Git, it's super intuitive, and it allows you to give the candidates great flexibility for choosing when to work on a task, but at the same time, you can specify a time limit for how long they can actively work on it. The UI and the look and feel are so slick. And yeah, I will report on the candidates experience in a bit. But for now, if you're also in need of a take-home assignment platform to streamline your hiring process, please have a look at https://codesubmit.io
They also offer a free trial, but now back to Nader. Nader, is a web and mobile developer who specializes in building cross-platform and cloud-enabled applications. Right now, he works at Amazon Web Services where he develops features in the client team and improves developer experiences for client side SDKs. Before he founded his own training company, specializing on React Native and trained engineers from organizations such as Microsoft, Amazon, the US Army, and many, many more. If you work with technologies like AWS or serverless, you definitely came across his writing, his videos or even his book. So, I'm super excited to have Nader here with me today. Nader, welcome to the show.

Nader: [00:02:00] Thank you. Thank you so much for the great introduction and thank you for having me. I'm excited to be here.

Michaela: [00:02:05] I'm really happy that you're here. So, Nader wherever I look, be it on Twitter, GitHub, YouTube, Dev.to, you are super active on all of those platforms and also you have a huge following of each of them. How did that start and were there also times that you were writing blog posts or making videos or implementing some stuff and hardly anybody saw it, or how did you break out of the unknown? How was that for you?

Nader:[00:02:30] Yeah, absolutely. So, I mean, I really started my whole social media and I guess outbound presence about four and a half years ago. I think I would say. I was working at a company called SchoolStatus and it's a small startup here in Mississippi where I still live at the moment. I was starting to become very active in the community in the sense that I was following people, I wasn't really active myself, but I was getting to know the people in the community and I was starting to go to conferences and I did run a meetup for a little bit. So, a lot of the people that I was really admiring were often creating blog posts and videos and stuff like that. And I always wanted to get into that. So I think the first blog post I ever wrote was a introduction to webpack. And the reason that I wrote it, I was waiting to find something interesting to write about, but I always felt like my ideas were not interesting and that someone else knew something more than me, still get that feeling now. But anyway, that's a different subject. But basically, when I was trying to build out a webpack configuration, I couldn't find really good documentation on how to do it. So I started, you know, piecing together things that I found, and I ended up building out my webpack configuration and building my app. And I thought, you know what, this would be really helpful for people. I think, because if I had found this exact writeup, it would have helped me a lot. So let me go ahead and create that. So that was my first blog post ever. And it actually, really encouraged me to write a lot more after that, because it just happened to do really well. I wrote on Medium when Medium was still one of the most popular places, I guess, for developers to write. Now I think people are, all over the place, but it did pretty well. And it was very encouraging and yes, after that, I've had posts that get zero to a hundred views. So it's hit or miss, even if you have a massive following, if you have 10 times as many followers as I do, or if you have 10 followers, a lot of times it doesn't really matter. I think the content itself matters the most if you're creating something that is going to help people. Of course, having like a lot of people to share with helps boost that at first. But I think over time it will do well, regardless if you kind of create something that is really unique and interesting and actually is helpful.

Michaela: [00:04:47] Yeah. Yeah, I definitely think so. I mean one of the things that I, I'm learning Python since almost two years now, and I'm coming from a java or a C-sharp background, so very object oriented and a different flavor and different style of thinking. And I'm actually really having a hard time finding resources for Python that help me grow, right. Either it's really basic and I know the syntax and know how to do that but I'm missing how to do the architecture of a Python application. And I've built large Python applications already, but there must be better ways to do this. And so, right now, for example, I'm tinkering around with this idea, maybe I should write some of those resources, because I feel they are not there. But it's really scary to write that if there is nothing that you can fall back and say there is no true or wrong. Here we are going a little bit away from "Does it work or doesn't it work?" because how you structure applications is a little bit up to you. There's no, "This is how we have to do it". But how is that for you? If you're developing something or writing something that hasn't been done or hasn't been done in that way? I feel there's a little bit of risk. And for me, it holds me a little bit back to you know, feel confident enough that, well, I know enough of Python and I made enough applications and architecture that I have something to say. And I have maybe something to say for people that came from a Java background, this niche of people. But how is that for you?

Nader:[00:06:26] Yeah, absolutely. Even today I still get that feeling, but I think the interesting thing about technology is that it's not like math, where there is a right and wrong answer. A lot of times there are just best practices, but a lot of times you can see someone that's extremely successful or two people that are extremely successful. They can both have completely different opinions. And when you realize that it starts opening the door to, in my opinion, have a little more confidence and to be able put yourself out there and share your ideas. And sometimes you'll be completely wrong. I've created posts and I've said things in the past based on what I thought was right. And I've had experts chime in and say, "Hey, you're completely wrong". And I think that it's not the greatest feeling when you're trying to teach other people when something like that happens, but it's also a good feeling because you can then go back and update what you wrote and have that best practice there. And then over time we were just learning more. But I think one of the hardest things about creating anything is the fear of being criticized. And some of that criticism could be us being afraid that we say the wrong thing. And I think being able to feel comfortable being wrong is a big barrier to overcome, to be able to be successful. Like if you're a quote unquote content creator. So for me it was hard overcome that at first. And I think you'll still feel it, even as you become more and more comfortable, but it will still linger I think.

Michaela: [00:07:56] Yeah, I think so too. And I think it really has to do with this humbleness. And maybe what I try often to do is to have this learning perspective in my writing, in my things that I create. And now I started streaming on Twitch and it's the same thing here. I was really afraid, but I really try to go into that with this learning angle, where I say, "Oh, I'm new here, I really don't know what I'm doing". And this somehow lowers the barrier a little bit for me, or at least I feel like more confident to share the things because I set the stage for "Well, I'm not an expert here right, I'm just sharing my story if you want to listen. And if you think I have, you know something interesting to say, I'm happy, but obviously, you know, feel free (laugh) to do something else". So you were talking about being a teacher and I think this goes maybe a little bit back to this part of your career where you actually were a trainer right? So you trained engineers at large organizations and probably also small organizations. So many engineers for, for example, React Native. How, how did that happen? How did you start this company, this training company? And how did that actually go?

Nader: [00:09:07] Yes. So I did training and consulting for about year and a half. And I was in between you know, it happened when I ran that company in between where I'm now, where I was before. So I was at a startup as an engineer, today I'm a senior developer advocate at AWS. And in between these two jobs, I was running my own company and I actually still have that company. And I still have a little bit of revenue coming in from contractors that kinda of, that get leads through that company. But it kind of happened organically I would say. The whole, you know, the whole training industry to me is extremely, extremely interesting. And it's changed a lot, I think, of course, since COVID happened because a lot of the higher priced trainings were like onsite trainings, where you would come and do you know, work with the team on site. But I do know of some companies that are actually still able to do some pretty high value trainings remotely and still get a decent amount of money for those trainings. But the way that I kind of got into it was, the React Native ecosystem is kind of what I was specializing in. And when React Native first came out, I was one of, or my company was one of the first companies to actually build and deploy a React Native app to the app store. And I was fortunate enough to be on the team that built that or actually I was the main developer behind that app. So in the first few months after React Native came out, I was, you know learning React Native and I was all into the documentation, but also on Stack Overflow, asking questions and also answering questions. And I think that whole process of getting in really early on that technology and learning it and helping other people with it, it kind of helped lay the groundwork for me to kind of brand myself as a specialist. And I think when you brand yourself as a specialist in any technology; this is kind of what I learned during this whole process. It's much easier to kind of grow and create a training slash consulting company. So when I was doing all of this work with React Native, I built a reputation on Stack Overflow for answering questions, and I became one of the top people there within that tech and what ends up happening is actual recruiters, companies, and people go on to Stack Overflow to actually find people to work with. So I started getting contacted by a lot of different people within the industry asking me to do consulting, just organically. And the consulting that was coming in was actually paying a lot more than my full-time job. I was making, you know, a very low end developer salary actually at the time, because I was working at a startup in Mississippi. It's nothing like a salary, of East or West coast, especially. So, when I started doing this consulting a little bit, I started realizing, okay, I'm going to make a lot more money doing this than working for the startup. So I was like, I'm going to really try to make this thing happen. So I built a website for React Native training because I had a couple of training gigs that came to me and I actually would take my vacation days to go do these trainings. And they were, I would make more money in one or two days than I would in the entire month or two. So I was, you know, really wanting this to happen. And I was okay. So I built this website called React Native training. It was just a single page written with jQuery and HTML and CSS, nothing fancy. And I started just getting a ton of leads coming in from that and landed a really big contract. That was the one big contract that kind of enabled me to quit my job. So I took that contract, quit my job, and I did that full-time for about a year, maybe year and a half. And it was really exciting time in my career because I was really, that was kind of when I moved, I would say from being kind of unknown in the community to kind of being a little bit more well known just because I was able to network with so many people and focus a lot of my time on building, you know, relationships and stuff like that. So during that year, year and a half, I did nothing but training consulting and grew my following on Twitter and stuff like that a little bit and, you know, learned a lot about consulting and it was a really interesting learning experience because I learned that there are things that I really liked about, about it. And there are things that I really disliked about it, but I think the one thing that I learned is that I didn't know what I was doing at all. And if I had kind of known, you know how to run a business, I would've maybe stayed doing it and maybe been a little happier. But by the end of that year, I was just kind of miserable. I had as much work as I wanted coming in, but I was overworking myself to the point where I was working like seven days a week, traveling all the time and the opportunity to work with AWS came around and I was able to kind of like look at it as a way to kinf of take my foot off the gas a little bit, and also learn new technologies and almost kind of get paid to learn those because, at AWS, I'm learning so much, but I'm still getting paid. Whereas when you're training and consulting, you're expected to be the specialist and you can't really take your time to learn anything new that's outside of that, because then you kind of are almost taking money out of your own pocket and it starts becoming this whole psychological thing that messes with you. At least it did for me. So that was kind of my story, I guess.

Michaela: [00:14:14] Okay. Yeah. That sounds really interesting. And so, you were saying some of the things that you really liked and some of the things that you did not like so much. So what are some of the things that you thought you're not doing so well or that you didn't like so much? I mean you were touching a little bit on it with this you know overworking and chasing a little bit the money probably as well right. But what are the things that you think if you would do it again, that you would have to you know, level up as a businessman, maybe.

Nader: [00:14:45] Yeah, I think that the number one thing I did not do well was scaled the business and the number of leads coming in. I think that when I first started, I had no idea around the demand that there was for that type of work. And how, if you kind of positioned yourself the right way that you can funnel almost all of the leads for people looking for a certain thing, because, you know with SEO and a combination of, writing, I read a book called "React Native in Action", I was able to kind of get a lot of leads for this, you know speciality. But I ended up, you know, I could only work so much and I didn't really know how to hire people and do stuff. I did work with a couple of contractors, but I didn't do it right and I ended up not really having anyone that, that I could consistently rely on. So I ended up turning away a lot of business and, I was also not that good at managing the different contracts that I had. So, I would do a training. I would do it. I would, I would also take on some hourly consulting and then sometimes one of those consulting contracts would just completely, you know, drop. They would be finished and they wouldn't need me anymore. And then I would have a massive spike in income and then my income would go down a lot and that would kind of stress me out. But I think the main thing that I learned was that I needed to be more organized. I need to learn how to work better with people. And I need to learn how to keep up with my paperwork better because I made a lot of money, but I didn't pay my taxes. So the next year, I mean didn't like spend it, I still had it, but the next year I had a big tax bill. So a lot of the money that I thought I'd made, I actually ended up owing in taxes. And, if I'd actually maybe expensed my stuff better too. I understand now that I could have written a lot of things off that I didn't write off. So all types of stuff. I think you could probably look at every mistake in the book that someone could possibly make and I probably made (laugh) those, made those mistakes.

Michaela: [00:16:39] (laugh) Yeah. And, you know, I started my own consultancy and training businesses. It's also why I'm so interested in this, right? A year ago I went full-time, I did a little bit before also, as you did, but I was still employed at Microsoft. But a year ago, I quit my job and, you know started doing this training and consulting business full-time and I was also really afraid. Like I'm telling this story quite often, I was like "Oh, I can't do it right". I have to do MBA first to actually do it, right?. But then I don't have the time to do that. So I just do it. And I'm struggling with really similar things.
I'm struggling with sales because I don't want to do sales. I don't know how to do sales. So there's really a lot of procrastination going on. So everybody that finds me just finds me through organic stuff, right. So recommendations maybe, word of mouth, my blog and things like this, but I've never really reached out to anybody (laugh) and, you know, tried to sell them my services. And then the accounting part as well right, I have no idea about this. So I see a lot of similarities and I think a lot of engineers that are thinking about, you know making this start maybe have the same struggle. And yeah, it's really interesting. I think in one point I have grown in the last year, I've got better, but I also make hundreds of mistakes in all regards to the business aspects and everything. Every mistake that you can do probably I also made. And so then AWS came along and they wanted to work with you. They also were getting to know you through your social presence and your presence on Stack Overflow, or how did that you know, connection happen here?

Nader: [00:18:19] So I think the first connection happened on Twitter. I kind of connected or just met with, met someone based on a tweet that had gone out that they were kind of engaging with another company. And it was having to do with mobile development. And I was like, "Oh, I wanna, I wanna do this thing, but do it with React Native" because you know, that was my thing. And somehow, you know, we ended up DM-ing each other and meeting, and then a few months later I was going to be in Seattle, doing a training at Amazon anyway. So, the guy that reached out to me worked at AWS. So I was like, "Hey, I'm going to be in town" And he was like, "Okay, there's this like meetup so come hang out". So I decided to go to the meetup. And I hung out with them at the meetup. And I met who is now my manager. And they were telling me about the project that I'm working on now. And they were saying that they were building out some React Native prototype, and it's going to be open source and "Hey, why don't you check it out?". So I was like, "Okay, cool". So I went back home. A few days later, I checked it out and I worked with the engineer on that team remotely and kind of just contributed their open source for free really, but also kind of you know, was learning the technology out there that they were going to be building out.
So during that process of helping them we ended up you know, building. I ended up adding, a pretty decent feature, I think, to what they were building out. And based on that relationship, they then reached out to me a few weeks later and they're like, "Hey, we're hiring for this position and we think you'd be a good fit".
And I actually was, I was like almost a hundred percent against it. I told my wife about it, I was like, "Yeah, this is cool, but you know, I'm working for myself and it's a little less money". Believe it or not, it was a lot of money, but it was less money than I was making. And I was like, "Yeah, I don't think I'm going to take it". And they were like "Look, just fly out to Seattle you know, have an interview. We want to tell you more about it. We think you'll like it". So I gave it some thought and you know, I started weighing everything and I was like, you know, thinking how it would be nice to actually be not held accountable for myself, to be able to have weekends off. All these things that I weren't doing and not have to rely on myself for a little bit. So I was like, "Okay, let me go out there". So I flew out to Seattle and I met with the team and the things that they were working on were like extremely, extremely interesting to me and really was like a perfect fit for kind of where I think I wanted my career to go anyway. Because I was really focusing on front-end and mobile development, but I didn't know a lot about back-end development and the things that they were building were exactly for people like me that were new to cloud computing. And they just were really, really interesting to me. So the opportunity to actually be on this team and a lot of the people that I met were just really, really smart and I'm always drawn to of course, work with those types of people. So I was like you know, "Okay, cool". So they gave me the offer and the offer was good. So I was like, "You know what, let's just do this". So I took the job and it's actually been probably one of the better decisions I've ever made, I think.

Michaela: [00:21:17] Yeah, it sounds really cool. And so when you were flying you in for this interview, was this more like a persuasion interview where they tried to convince you that you should start, or did you have until, after even writing you know, some major contributions in open source and your presence and the trainings, did you still have to prove yourself in the interview process? Or how did that, that go?

Nader: [00:21:41] Yeah, it was a real interview process, but I also got to kind of see the office and meet a few of the people on the team. And there was the typical five-step Amazon interview and there was a technical interview and I didn't do that good on the technical interview from what I understand, but I did do really well on all the other ones. And I did well enough, I guess, for them to kind of, you know roll the dice on me. But I did have the regular interview process, yeah.

Michaela: [00:22:05] Okay. Yeah, this is. I mean, this is really stunning for me. Like if they really want a person and then they still send them through this, well, probably very consolidated process, right. So nobody can go around it which, you know. Okay. So yeah, what's really interesting to me as well is, if you could tell me a little bit about serverless. Because this is your specialty, I understand. And you thought this is super interesting to you because you were not you know, familiar with the backend so much. You were really focusing on the front end. And so you said it's technology that enables you or front end developers right, to actually write full-stack applications. And this is also what I, I read some of your blog posts for this interview today right, to be ready. And this is also what I read in this blog post about full stack serverless, right. Where you have, where you have actually the infrastructure as code and you have like the backend actually as services that people can just consume. But if you have to explain it, what is serverless, who can use it and why is it so powerful?

Nader: [00:23:11] Yeah, so serverless and, and, you know, the early days I think was just basically the idea of functions as a service. And I think it's kind of grown now to be more of like a philosophy around how you build infrastructure and how you build services. And I think if you asked me, like today what the definition of serverless is I would say it kind of encompasses a couple of things. Uh, one is like not managing your own infrastructure. Two is paying for the compute time versus the compute power, and basically being able to rely on someone else for the infrastructure to always be running. And also the last thing is the ability to scale up and down without having to also manage that, that functionality yourself. So basically buying into services and anything, any APIs and things like that that are built and maintained by someone else, but that are still configurable enough for you to actually get the work done that you need for, for your app or your, you know, your company. So I think that what you're basically, the thing that's really interesting to me is like, you know, I talk about serverless a lot, but I think what you're seeing in the whole cloud computing industry is that a lot of the things that were considered cloud are kind of moving to be more into the serverless paradigm. So, so even if you look at kind of, re:invent happened this week, there were a lot more, you know, things that were, that were released that kind of enhanced serverless. So for instance, services like that, like Amazon Aurora, they, which is basically a Postgres or RDS relational database service, they're making that serverless. So, instead of having to kind of like scale up and scale this down yourself, you can just, you know, have an instance and everything scales up and down for you. So I think the idea of being able to kind of have a lot more you know, powerful services on the backend that you don't have to deal with yourself, that you can just kind of buy into whenever you need them and focus on mainly the business logic that kind of differentiates what you're building from everyone else. And when you kind of are able to focus less time on things that can be abstracted away, you're able to iterate faster. You're able to, to. Do things like experimentation, which will also set you set you apart. Because let's say you want to build out a new feature and you have two options. Either build everything from scratch, and it might take you four weeks or take advantage of some managed service and it's going to take you one week. The companies and the teams that are using the things that are more, that give you more velocity I think, are fairing better. And also the, if you, if you think about the engineering cost that goes behind, building things from scratch versus paying for something, you know, as the usage is there; it just makes a lot more sense I think for a lot of companies. It might not, it might not make sense for like a billion dollar company, you know, but if you're a startup or even like, you know, a 10 or 20 million dollar revenue company, being able to kind of experiment without having to pay for it until you hit some type of scale is actually, it just makes a lot of sense, you know. Especially as engineering costs continues to go up because developer cost is going up. So any way to kind of cut costs down by having less developers is always a good thing.

Michaela: [00:26:22] Yeah. I have so many thoughts to this topic, right. So one, one of them is I, I see such a connection with the no-code movement. And at the same time, right, I also see such a connection to this problem that we are having or facing, or maybe it's not a problem. Just the evolution that we are having in technology, that technology becomes more and more complex, right. And we were talking about this being an expert. You said, well, I was an expert. I was specializing in React Native. Nowadays, you cannot be an expert in everything anymore, right. So you have two bigger paddles and say, well, I know, I know, React Native really well, or I know View really well, or, you know, I know Python really well. Or maybe Python with some frameworks. But it's very, I mean, there are I don't know anybody anymore, right, that has several frameworks and then even, you know, have technologies and, you know, uh Coopernetics and whatnot, right. That you have this knowledge for everything. And so I think serverless here, it becomes also really powerful because you have some parts that are developed by somebody that is specializing or knows exactly their stuff, right. Their technologies, what's needed for, for example, the service that you're, that you need. And you can somehow outsource that. Then you can concentrate on the important stuff that you have to develop. And also train your people internally so that they have enough knowledge to really, you know, build a scalable, maintainable solution. What are your thoughts here?

Nader: [00:27:52] Yeah, I think it makes a lot of sense in that comparison. I mean, what, what essentially serverless is, is just another abstruction, you know, on top of things that we've been doing. And it's kind of just a name that people are throwing out there and it actually works in my opinion, people criticize calling this or that serverless. But I think when you start thinking of it less as like a thing and more of like a philosophy, the serverless philosophy, or maybe even kind of, a variance of serverless. So you can be more or less serverless. So if your, if your infrastructure is completely managed, then you're completely serverless. But if you use maybe servers that you need certain things that you have to deal with yourself, then you're less serverless. I think, you know, what you're seeing is a lot of companies that, like I mentioned earlier are having to deal with constraints like monetary constraints. Like they don't have a ton of money. So when you're working with a serverless, you're able to iterate faster, have less engineers and, and, come to market faster and just be able to build quicker. So you're able to, to do that with less money. And also when you're having to compete with other companies that are maybe in the same space as you, you have this idea of opportunity costs. So being able to come to market quicker, serverless kind of allows up and, and things like that.

Michaela: [00:29:08] So, if you would allow, let's say we are really starting from scratch here, right? We are a young company. We have like an idea often of an, of a service or business that we want to develop. We want to develop our MVP. How does serverless shape our architecture of this MVP? How would we go about, you know, understanding what should be, you know, a service, what should we, you know, keep in-house? Can we just, you know, outsource everything? I mean, some of the things that come to my mind definitely is authentication, right (laugh), which is complex (laugh). And we could somehow, you know, get that out of the door. But what are other things, the examples that we would get in and how would we even tackle this problem of developming MVP, if you want to use serverless technologies?

Nader: [00:29:49] Yeah, I think authentication is actually like a really great example because if you think about it, when you're working with something from scratch, you're, you're then competing essentially against these services that are also out there. So we have with something like Auth0 or Octa or Amazon Cognito, you have teams of people, specialized engineers that have spent years focusing on this single problem case, the single problem area. And they've spent all this time tackling all these edge cases. Do you think that your team can build something that will compete with them? And if so, maybe there, there is a, situations like that because maybe you have some really, you know, specific thing that's happening in your system that, that doesn't maybe work well with these, but for the most part, when you look at it like that, you can kind of buy all of this expertise, like as a service. And also it's usually more secure when you're thinking of things like that. But the way I would look at like a new app, is, you know, I would kind of typically say, okay, let's list out all of the functionality that we need. And then we take that functionality and we kind of say, okay, how can we make this happen with managed services or with several of our functions as a service with some type of database backends. And we'll take all of those features and we'll plug those all there. And if there's anything left over, then that's kind of where you would, then manage that yourself. And I think like by your, it's essentially a server-less first approach. So anything that can be serverless, try to make it serverless, try to fit that into either a managed service or build it using functions as a service with some type of database persistence layer, and then whatever's left over then that would be what you end up managing yourself. And I think that's kind of a good, a good way to go about it.

Michaela: [00:31:32] And so if we are having those services, let's say for example, authentication is, I think it's a really good example, but on the other hand is like maybe overused example. Can you give me one other example of what we could, what we could, for example, use it as a serverless service.

Nader: [00:31:47] Yeah absolutely. I think a really big one is the, well there's two really big ones actually. I think storage is, is, I mean, Amazon S3 is just massively, massively used and adopted, and that is a service. It's just a place that you can drop a video. It's like a file service almost right. You can drop videos, you can drop images, PDFs, and you can retrieve them. And it's a lot faster and a lot more efficient, a lot more secure than making these, in turning these files into binary and trying to store them in your database right. And then there's all of these. Security systems that are built around it, that kind of make it, make it even you know, more secure than something that most people could build. So I think storage, file storage with S3 and other similar services, you know, outside of AWS are good, a good, you know, a good example. And then the other are API services. So the two biggest ones at AWS are Amazon API gateway and AWS app sync. But there's others, you know, think of all these other managed services like Sera. Like, you know, I there's a million of them. I can't really like, I guess Firebase could even be kind of considered that, right? These are like API slash data services where you have like an API layer that is managed by someone else. But all the traffic coming in to that API layer can then be redirected wherever you'd like it to be. So this request goes to this database. This request goes to this function. This request goes to the CC two instance, whatever. And the API services are managed for you. And the only thing you're dealing with is actually the business logic.

Michaela: [00:33:15] Yeah. And so how would you, how would you implement that? Would you just implement it directly or would you have like, abstraction layer in between so that you actually can change services or, or, you know, serverless functions or something like that?

Nader: [00:33:29] Yeah, yeah. I think like when you're buying into these types of things, then you're, you, you are somewhat, you know, tied into or locked into those. So I think it, it's kind of, you know, when people are typically building in this way, they are, I guess, weighing the different trade-offs that they have. So I think when you're, when you're doing something like this, you're some somewhat tied into that, to that system. So for instance, I think a lot of, new customers are using stuff, or they're at least evaluating stuff like GraphQL with app sync, because a lot of the apps that are modern or basically, you know, interacting with a bunch of different clients. So you don't just have a web app now. You typically have a web app and a mobile app, right. But you don't want to build separate end points for each one of those different apps. So using graph QL, you can have a single endpoint and for, you know, your mobile devices, you might want to fetch a lot less data for, for you know, to have better performance, but on your, your web app, you, you know, you can assume that they have a better connection and more power. So you might fetch more data plus they have a larger screen. With GraphQL, you can just have like the single entry point and be able to just ask for exactly the data you need and things like that. So.

Michaela: [00:34:41] Yeah. And it makes sense. Yeah. And so you, you're writing books like you don't have like one book you have, like at least two, maybe you have more, I dunno, but the two that I have seen, how are you going about that? And especially if now staying with serverless and for example, full stack serverless, right? How, how do you go from idea to have this finished book and yeah.

Nader: [00:35:06] So I have two books. One is react native and action from Manning publications. And the other is full stack serverless from O'Reilly. And I think that it was, the first book was probably the most challenging thing I've ever done, you know, really, I would say in my career. As far as like a long-term project, it took over a year and a half to write that book because of a couple of reasons. One is the publisher I was working with, the editor that I had within that publisher changed like three or four times. And every time a new editor would come in, it took a while for them to get up and running and then they would go through my book and then they would give me all these changes and then I'll make the changes and then react native changes a lot. So there would be, then I would have to go back and make more changes. And then a new editor would go. It was just, it was kind of a mess, but it was mainly because of the rapidly changing technology that is react native. So. But that process kind of was really, you know, it was really, it taught me a lot, I think. Because when you're doing something for that long and you know, you stick with it and you finally finish it, it feels really good. And then my next book with, with O'Reilly publications was a lot easier because I was able to actually write it a lot faster and get it out there. And then also they have a way for me to update it using, uh, git. Which is, you know, what I'm very familiar with. So anytime something changes, we create a new version, and we, and we have it ready to publish. But I think that, I think that, you know, when you have an idea for a book, there are just so many options now to publish, right? You can go with one of these publishers like I did, or you can self publish. And I think there's trade-offs to each. I think the reason that I chose to write my first book with react native was because I was not really known at all and I wanted to kind of have a book, and have it published by a known publisher because that would give me a little more credibility. And I think that those are the types of questions and trade-offs people have to consider. I mean, once you get a lot of credibility, you can self publish a lot easier I think, because people already know you. But if you're kind of brand new to the community and people don't really know you, it's, it is a nice way to earn trust when you're working with one of these nine publishers. But it's just a lot of, uh, a lot of work long long-term, you know, writing. But, but a lot of people, you know, find it very fulfilling and you know, people, people are putting out some really interesting books lately out there I've seen.

Michaela: [00:37:24] Yeah. And why did you choose for the second one? I mean, you had already a following. You were known. Why did you choose to go again with a publisher?

Nader: [00:37:32] So O'Reilly publications is, you know, O'Reilly is kind of what I learned almost like if I read any technical book learning what I know now, I would say about 80% of them were O'Reilly. So having the opportunity to actually write one was just a really cool opportunity. And I just wanted to do it mainly because, I have so much respect for O'Reilly and I had actually heard a lot of good things about the writing process. So I just did it because you know I thought, I thought that was really cool. It also, it kind of feel like the reach that it would get might be a little better with the whole platform that they own. So, yeah. But it could have gone either way, I think for my second book.

Michaela: [00:38:11] Yeah, I can, I can see that. And so did it help, did the editor help in the whole process? Did you feel that it's actually a better book because you had like an editor that was working with you on the book?

Nader: [00:38:23] Yeah, the editing process was really, really seamless and it was cool because I was writing with, you know, putting my, my pages in GitHub, a git provider like GitHub. So anytime that they would make changes, I could just run a git pull and pull down the updates. And most of the time they were doing really tedious stuff that I didn't really want to do anyway, like grammatical errors and formatting, which was really nice. And they did, they did a lot of that. It was pretty painless.

Michaela: [00:38:48] Yeah, that's really cool. So one thing that is still in my, on my mind is, um, we talked a little bit about when you have like serverless and I'm thinking is a little bit like the no-code movement as well, right? Where you have like this abstraction layer and you were also saying, well authentication, right. People know so much about authentication. So if you want to do a similar service, how much would that even cost you to, to replicate, right; to do, to maintain and everything. So, and one thing that you also said is like code is liability, which is, I really think this is so on point as well, but what's actually different? Is it, is it actually the same? Is the no code movement also serverless? Or is that something else and what makes it different from the serverless movement?

Nader: [00:39:35] Yeah, that's a good point. I think they all kind of start overlapping, you know, here and there. I think that the, the, the no-code movement is really interesting, but I, I would say that you could kind of compare serverless in my opinion, closer to like the low code movement or like the low code tooling, because what serverless does is it kind of gets rid of automated things that you're doing a lot of times. So for instance, I mentioned the API layer. You know, you have, oh, with backend developers. A lot of times their role is to do a lot of things, but one of those things is often to just create API end points and those API end points like interact with some database. And I think what you're seeing with a lot of these managed backend providers is that they're kind of doing you know, that work for you so you can kind of like all flow it down and then you can give the backend developers more important stuff to do. But with the no code movement, you know, I think that when no code succeeds at something, it kind of, it just crushes everything. And it just does really well. But, but, but it's hard to do certain things with the no code. So for instance, before we had Wix and before we had these types of tools, you know, people that didn't know how to code; there was pretty much no way for them to create a website. Now they can do that. But I think for someone to go from creating like a marketing website to creating like an app that actually has big core business logic, that step seems to be a lot harder to do. And I think that's where serverless and no-code shine because they automate away, a lot of the tedious stuff and they allow you to kind of just get to the core business logic that is kind of differentiating what you want to do versus everyone else.

Michaela: [00:41:11] Yeah. I, I, it's definitely the, also like the granularity, right? So with the serverless, you have like really smaller building blocks that you plug in together. And as you say, it's, it's actually not no code because you definitely have to code and it's not even low code, right. (laugh) It's probably, it's really code, I don't know. (laugh)

Nader: [00:41:28] It's lower code maybe, yeah. (laugh)

Michaela: [00:41:30] Yeah. (laugh) I don't know where to distinction here comes and what it's actually called, right. A lot of things, even Excel is somehow coding, right? But Yeah. And, and yeah, this is how I see it, right. It's like, it's this. And Wix and all of that are really like these large things, but they're like smaller services. Now we have like Sopier for example, or, you know, like where you can have this automation off of processes. And I think people are really doing interesting. I, I'm not really too much into this no-code movement because I think if you can code (laugh), this is maybe the default that you do, but I try to, I try to get a little bit into it because I feel that I don't want to replicate everything, right. And spend my time doing things that I could actually take out of the box. So this is why I'm really interested in this no-code. And I also think that serverless is really, really interesting because you can save yourself so much time and even, I mean, I'm, I'm a solopreneur, right? So it means that every time I'm touching something, I really have to think, well, how I'm going to spend my time and how I'm gonna scale. And so I think those things can really, can really help here. Yeah. So one of the question that comes to my mind, and especially if I watch some of your YouTube videos is I feel also the complexity right now with serverless and those technologies in general, with our ever evolving technology base comes a little bit from programming to configuration, right? So even those

Nader: [00:42::57] Oh yeah, that's a really good topic. (laugh)

Michaela: [00:42:59] So like you have to open up your account online, right. And you're just installing something. And I'm sure I hit enter and there will come an error. (laugh) And I'm like, Oh my God, what I'm gonna do? Something is wrong. So I actually struggled quite a bit with setting up my environments and all of that. I feel it much easier to open up, you know, like just a very plain editor and start writing code. So I feel that a lot of this complexity that we first had in, in writing code is now in configuring your environment, knowing how to, you know, install and having systems talk to each other. How do you see that?

Nader: [00:43:36] Yeah, I think, I think it's a super interesting topic. And I think it's one of the things that we take for granted as developers, as we start getting more experienced than we're doing this longer, because a lot of the things that these abstructions start doing for us were things that we kind of understood anyway, but we're just kind of happy that we don't have to do it anymore. But a lot of times, those things, you know, need to be tweaked, or you have to kind of understand that those things are happening. And then when a new developer comes in, someone that's just starting, then they run into these issues and they don't understand like why they're running into them. And a lot of times it's because there has been something that's abstructed away from them that they don't really see. Or there's nuance, and that there's so much nuance that, that is there as you become more familiar with how the computers you know, work, the command line works, whatever your environment that you're working in works. And yeah, and, and, and all of these configurations are kind of like, you know, where those things can open up to give you an idea of like, what's going. And then, and then when you start thinking about, and this is slightly veering off, so I'm sorry. But when you start thinking about what we were talking about earlier with cloud computing and serverless. What you're also seeing is kind of the ability to provision infrastructure based on infrastructure as code, which is essentially kind of like configuration. And, I think what's really going to be happening over the next few years is that people are going to be building a lot more powerful tools on top of infrastructure as code and abstracting that away. So I think the tools are going to be things like, like seek command line interfaces, but also GUI, so user interfaces that allow you to kind of create infrastructure as code under the hood. One of the really interesting ones that is out there it's called Stackery. They, they allow you to kind of, drag and drop. And then under the hood, they're basically creating infrastructure for you. And then the amplify CLI, which we work on does the same thing. And then there's CDK, which kind of does the same thing, but it all, yeah, it all comes down to like configuration and when you're, and when, and if you're a new developer coming into it and you're using the tools like we create, and you don't really know what's going on under there, there's only a certain point that you can get to before you need to kind of understand what is happening. So I think there's a big trade off there. Like, yes, it's great for new developers, but it's also kind of like a possible area where they can, they can, you know, hurt themselves or, or they can just be limited because they're going to run into some issue and they're not going to know why that issue is there. Whereas someone like us that's been doing it for longer. It may take it for granted, unfortunately.

Michaela: [00:46:05] Yeah. I mean, for me, it's always like, uh, back early days when I did my bachelor's right. It was, it was databases. You had to install like a database and you have to really make sure that everything runs. And it was like, Oh my God, is that gonna explode? And it's, it's like a black box, right? So obviously you can figure out, but it takes a lot of time. And, and now I have the same feeling here. Every time I'm using a new tool or a new API or something that comes out of the box and that's actually a black box, I always feel like, Oh, is that, is that gonna work now? And if it does, I'm like, wow, (laugh) this was amazing. But very often it doesn't, right. And, and as you said, and it's really hard, I think to troubleshoot, especially if you're new to it, because it's a black box and so how, how should I know what's going on here right now? So, yeah, there's definitely a trade-off and I, I'm really excited about what you were saying about these tools that we'll build on top of that. So how do you envision that will be, will that be extensions of the IDE or will that be something separate? You're talking about the AWS CLI, right? So this is probably separate from, from the IDE, but could it be also part of it or how do you think, which direction will that all go?

Nader: [00:47:22] So I think they just actually based on some very recent experience, because we just released actually something on my team this week that kind of gave us a very good idea around what people like. But I mean, what, when, when you go, for instance, you log into something like Azure or GCP or AWS and you start creating things in the console, you're basically provisioning infrastructure, but you're having to do it, you know, in this, this user interface. But you're only really able to kind of like do one thing, right? If you want to kind of replicate that there is no way to do that. And when you're thinking about like real-world production systems, you typically have different environments and you have built processes and you have testing and all this other stuff. So you typically need some way to kind of automate that. And that's kind of like, you know, where infrastructure is code comes. But when you're like a new developer or even an experienced developer, you start looking at some of the older ways of writing infrastructure's code. If you're using AWS, the one that you might think of is cloud formation. Just to create something pretty basic, it's just a lot of boiler plate and there's a lot of, you know, knowledge that you have to have to kind of make that happen too. Because you have to know which keys and which values do what, and you have to kind of know what options there are. And just for a single service, it might take you dozens or even hundreds of lines of code with a lot of things. So that's just one thing. Imagine putting, you know, five or ten things together, or a hundred things together, which you know, is not uncommon. So what basically all these newer tools are kind of doing is they're kind of, they're, they're giving a way to kind of go back and forth. So before you either had the option to do in, in, in the console, but it was harder to replicate or you could do it in this configuration file, which was hard to get started with. But what if there was a visual way to do it that then exported that into a configuration file that you could then run in some CICB process or just hand off to a new developer and have them copy it. I think that's where we're headed. And that's what the amplified CLI does. That's what kind of, what CDK? Well, I wouldn't say that's what CDK does. CDK is just another way to kind of do this where you can write TypeScript or Python, but I think in the future, what we're going to basically see is, these graphical user interfaces that create infrastructure's code that can export, you can modify them and then redeploy them. And then they create that create the infrastructure. So you have kind of this back and forth going on and it, and it will, it will vastly lower the barrier to entry for people, for cloud computing, because it will allow them to build out these real world production systems without having to know a million things, you know.

Michaela: [00:49:58] Yeah. Yeah, I can see that. So actually I took a lot of time from you already, and I think this brings us somehow also to the end of this interview. Is there something that you think you want to give my listeners on the way, especially if they are interested in serverless, for example, or cloud computing, mobile development?

Nader: [00:50:20] Um, there are a lot of things I would probably like to plug. So let me think about this for a second. I mean, if you're interested, you know, in, in cloud computing in general, I would definitely just recommend checking out all of the different options that are out there. There's AWS, but there's also Google and there's the Azure and they all kind of have their own, you know, communities and they're all pretty cool. I would probably, you know, I would just encourage to find different people within all of those communities and follow them and just look at the stuff that they're putting out like creators, you know, people making videos and stuff. I don't want to plug anyone in particular because I'm kind of a big fan of a lot of different things and it's hard for me to just pick one or the other. But I would say, even though I work for AWS, I encourage people to look at, look at everything that's out there and find what is, you know, not only interesting to them, but kind of what's easy for them or what makes sense to them because I think they all are done a little differently and some click better for others than, than others.

Michaela: [00:51:20] Yeah. Yeah. Maybe I have a little follow up question when you're saying it that way. Do you think that you can, let's go back to this example that I had; we are building our MVP or are we designing our system. Can we take serverlesss from here and from there and from over here, or should you stick with one provider with one company that's providing those services to you?

Nader: [00:51:43] I just, I always say pick the best tool for the job, whatever that is. So for you, if that's all with a single cloud provider, then that's great. If, if you want to kind of expand and do multiple cloud providers, then if that makes sense for you, then do that. I think I'll always, and this is slightly against the messaging I think that you, that you hear a lot of times from cloud companies where they kind of say, you know, multicloud is what they call it. And they say like, don't do it. But I'm just kind of the type of person that says do the best tool for the job. There's there's someone that I follow. His name is Joe Emison and he's actually built and sold multiple companies using serverless technologies. I think he's on his third company. Wildly successful, super smart. Um, he, he uses AWS, but he also uses other cloud providers for different things that, that, that are done really well, that, that you can't really find within AWS. So I'll say find the best tool for the job, whatever that is. So that's kind of my take on that.

Michaela: [00:52:40] Okay, yeah. Thank you so much Nader for being on my show. It was really a pleasure to talk with you about all these different topics and, and hear you talking about so many things that you experienced from, you know, being an entrepreneur, to working at AWS and author and blogger, tutor, everything. Thank you so much for being here.

Nader: [00:52:59] Thank you, Michela. It was really, really nice to be here. I appreciate you having me on.

Michaela: [00:53:03] Yeah, thank you so much. Okay. Bye bye.

Nader: [00:53::06] Bye bye.

Michaela: [00:53:07] I hope you enjoyed another episode of the software engineering unlocked podcast. Don't forget to subscribe. And I'll talk to you again in two weeks. Bye. (outro music)

Copyright 2022 Doctor McKayla