We're talking AWS on Discord! Join us

Episode 18

Austen Collins: Serverless Framework, Serverless Cloud, Serverless Ops Console, and the Future of Serverless

Austen joins Adam to discuss the evolution of Serverless, Inc, the many products they're building now including the Serverless Framework, Serverless Cloud, and Serverless Ops Console, as well as the future of serverless development and monitoring.

About Austen

Austen is an entrepreneur and software engineer located in Oakland, CA. His specific focus is on building cheap, scalable Node.js applications while minimizing DevOps requirements as much as possible. An enthusiastic AWS Lambda user from day one, Austen founded the Serverless Framework (formerly JAWS), an open source project and module ecosystem to help everyone build applications exclusively on Lambda, without the hassle and costs required by servers. AWS Lambda can significantly reduce the total cost of ownership of all software projects. Every ambitious developer and company should take advantage of Lambda, and Austen will help you get started.

You can find him online on Twitter, LinkedIn, and GitHub.


Adam Elmore: Hey, everyone. Welcome to AWS FM, a live audio show with guests from around the AWS community. I'm your host, Adam Elmore and today, I'm joined by Austen Collins. Hi, Austen.

Austen Collins: Hey, Adam. Thanks for having me.

Adam Elmore: Yeah, thanks for joining. So, I want to start with your story and really specifically your story. So, your background in tech and then leading into, we're going to talk a lot about Serverless, Inc. obviously, but Austen's story and career.

Austen Collins: Sure. Well, I describe myself as actually a product person before an engineer. So, always been an entrepreneur, always been obsessed with just creating great products that make people happy. I grew up in Southern California and have unconventional background. I'm a self-taught programmer and was around a lot of the film industry when I was younger. So, I spent a lot of time in Hollywood, actually. And my first company was a multimedia company that did post-production for film and television projects as well as all the web assets. So, we were this one-stop shop for a whole bunch of whatever, whenever people needed to go to market with their film or their television program, we would create all that stuff for them.

Austen Collins: But in Hollywood, I got an education on a different type of engineering and that's an engineering for emotional impact, which is a science. And it is just like punching up a script is like punching up code. You're engineering it for the right effect. And I think that fed into me just being more of a very product-focused person overall in engineer. And at some point, I think it was 2009, I left Hollywood and moved up to the Bay Area just because reality was getting stranger than fiction. So, with all the stuff going on in tech, you've got cryptocurrency stuff, drones, state-sponsored hacking. It was just so irresistible and the startup scene and the innovation that was happening. And Hollywood was seeming very formulaic in comparison to what was happening in the Bay Area and in the tech realm. So, moved up there and started just focusing on startups.

Austen Collins: Again, I grew up with a family of entrepreneurs. Everybody was just like just start doing. And so, moved up to the Bay Area to do startups. And was working on a startup when I started playing around with this idea, which eventually came the Serverless, the Serverless framework. I think it was 2014 when they announced Lambda and I'd always been an AWS fan. They're working with AWS since the beginning and when they announced Lambda, it just, it spoke to me like nothing else had for a long time.

Austen Collins: And I think specifically it was because it seemed like technology could finally get out of the way, which is something I've been looking for since I started. Because if you're really obsessed about product and getting things to market and building things that people actually like, there's just so many timeless hard problems that you need to be focused on 100% of the time product like market fit. How do you build something that people truly love and use all the time? That is just the timeless hard problem. And technology is both an enabler that gives you so much leverage to solve that problem, but it's also such a big distraction. It's an incredible distraction.

Austen Collins: I think that's one of the biggest things that the challenges that engineers face on a daily basis. It's probably been that way forever, but as now, I'm a CEO and I manage, teams of engineers and whatnot. It's always, we're talking about the latest way to manage state and react or whatever came out on Hacker News this morning. And so much of the time gets sucked up talking about this stuff. Meanwhile, those timeless problems of like how do we build something? We make developer tools at Serverless Sync now. So, I wish our conversations were just focused on that stuff all the time.

Austen Collins: So, when Lambda came out, I thought very much like this speaks to me as well as it's a strategic inflection point for the Cloud, as we know it. And that is this community wide acknowledgement that it's time for higher levels of abstraction that are focused on business problems, not lower level infrastructure concerns and we're going to make APIs for all the things. And composing these levels of abstraction to build applications is what the future Cloud is going to be all about. I started Serverless framework early 2015 and so, way back then there was no Serverless buzzword or category, anything like that.

Austen Collins: Lambda was pitched as a venture-driven compute glue code. And it was very speculative that there was something happening here and a lot of it, I just felt, personally, very alone on this journey, trying to use this stuff in the early days and thinking this is a disruptive event for Cloud. Something's changing. I couldn't quite tell what it was and it's not as formalized. It wasn't as formalized as it is now. And so, going through those early days, I saw the technology. I thought it was amazing, but of course it wasn't accessible for developers. Lambda in the early days of AWS and API gateway had no docs.

Austen Collins: It was to write the first version of the Serverless framework, I had to go into the AWS console and actually look at all the network requests that the API gateway dashboard was making to truly understand how to build a framework that uses a API gateway and a lot of these new services. It was just such a mess, but the promise was there, the potential was there. And so, while I was working at another startup, I wanted to build it all on Lambda, but just that developer friction and inaccessibility was there. So, I started creating this project, which eventually became the Serverless framework and launched that in mid-2015, put it out there. And it took off like nothing I've ever seen before.

Austen Collins: And like all product people worked on a lot of stuff that hasn't worked, but sometimes you find product market fit and you know it when you see it, because it's like you kicked over a rock in your backyard and all this oil just comes spraying out or you find a treasure chest or something. It's very visible when people like your stuff. And launched Serverless framework mid-2015, it was racially called JAWS. And it was, at that time, we were playing a lot with the Serverless buzzword and using it to be frank, almost as a little bit of like a troll.

Austen Collins: A troll in the effect like, "It's time for Serverless applications," a way to describe what we're doing. And we created this narrative around Serverless and put all this stuff on the GitHub ReadMe that was propaganda-ish like Salesforce, Marc Benioff-ish. We used to have no software and so, we had all these badges on the framework ReadMe that said, No servers guarantee, 100% server free." And we started created this narrative and I think this movement.

Austen Collins: And so, humble beginnings in Southern California to moving up to the Bay Area to do startups. And ultimately, I think a lot of my story is just somewhat stumbling into a major technology category. And that maybe one of the greatest corporations of all-time. Adopted as one of its key differentiators for their entire platform for the next 10 years. It's been quite an adventure.

Adam Elmore: Yeah. And so, it strikes me and I didn't really know, honestly, a lot of your story and I should. I should have done more homework probably, but it strikes me that you were the perfect person to build Serverless. The fact that you're product driven. That it really is about the end result and delivering that value quickly, you were the right person and you were at the right time. So, you had the market tailwinds just timed out so well. Yeah, so kudos to you.

Adam Elmore: And it never occurred to me, did Serverless, did the name Serverless come from Serverless, Inc.? Did you coin it?

Austen Collins: Did not, did not coin it. Definitely, not. It was always around one of the original folks at Iron.io. I think it was 10, had posted something on Serverless. It's a very old article, I think it was 2011. It was the first time I have seen it. The story around the buzzword, is at least from my perspective, is I had seen it before. The first time I saw it was, I think it was early 2015 when Tim Wagner, the original GM of Lambda, had put in a blog post, he was talking about Lambda and API gateway. And in the middle of the blog post, somewhere there, he said you could use these services to build entirely serverless applications.

Austen Collins: And it wasn't the headline of the blog post. It wasn't anything. It was just one adjective that was just stuck in there. And as soon as I saw that word, I loved it. Because obviously, it's not technically accurate and the literal truthers are always trying to come out and point out the inaccuracy of it. But for, again, my background was designing for emotional impact and thinking about that from the product perspective and whatnot. And when you say Serverless to a developer, you could see them light up as a metaphor for everything that is complicated, that is maintenance, that is not innovation.

Austen Collins: And I pay attention to, hey, at least I try to perceive what makes people lean forward it in their seat and light up and start nodding unexplainably. And Serverless was something that did it for me. And when I would tell everybody else this word, I see the same results. So, we came out with, it was originally called JAWS, the framework. And at this time there was not a Serverless buzzword. Again, Lambda had a totally different categories it was trying to break into. But in the JAWS re project in the ReadMe, it was JAWS, the monstrously scalable serverless application framework. And then all those badges were put on there.

Austen Collins: And I love design and visual storytelling, so the logo of JAWS was this big shark and it just felt like a blockbuster event. I wanted to make this project feel like something big was happening here. And to embody it with this mascot, this shark logo. And just have this cool visual storytelling merely on the GitHub, ReadMe was a fun way of getting a narrative out there in addition to what I hoped was great technology. And I'm a big fan. If you want to put out something that's really disruptive, I think you have to have a great solution, a great product. But if you could wrap that in the emotional charge of a story, that is the one-two punch that makes a big difference. That hopefully, you hope rallies peoples in the masses.

Austen Collins: So, went crazy, putting the Serverless buzzword all over the ReadMe and tried to create a category almost. And you made it sound like it was so well thought through. To be frank, it was not as intentional and structured as it sounded. It was very much like it just feels like this needs to happen and I can't explain it and I'm going to move in this direction. And that time, Serverless was laugh at, so it was not totally accepted. It was just people were like, "You're going to Serverless, what is this? This is so stupid." And, the first Hacker News comment when I posted JAWS on Hacker News was like, "This is a horrible idea." And-

Adam Elmore: It's so Hacker News.

Austen Collins: Yeah. That's how they greet you when you post everything on Hacker News. And so, it was just like in the early days, it was just filled with ambiguity and uncertainty and a lot of antagonism everywhere. It was not clear. It was just a rational behavior perhaps on my part to try and create something. And ultimately, turn this framework into a product that made this cool new architectural pattern accessible to developers and it made it exciting. And I also thought that shark was a very cool successor after Docker, the Docker whale.

Adam Elmore: Oh, nice.

Austen Collins: I had a lot of visual story time that I never got to, because of course, we ran into trademark issues with the name JAWS and a shark logo.

Adam Elmore: Oh, wow. Yeah.

Austen Collins: I should have seen that coming immediately. It was like posting a project. It was amazing viral moment and then a whole bunch of legal threats. And so, it was just like the good and the bad altogether. And so, eventually, as soon as-

Adam Elmore: That started lift.

Austen Collins: Yeah, as soon as we posted the project, it was like, "Oh, shoot. We're going to have to change this." But I had so much cool marketing stuff I want to do with that shark icon and I had in my original pitch deck for the company, the first slide was the Docker whale floating in the ocean upside down with a big bite out of it and an XO over its eyes.

Adam Elmore: That's so awesome.

Austen Collins: Yeah. It was just, I'm a big fan of GIFs. I'm always making them. And the slide was perfect. The whale was just floating upside down. Those were the first slide. And then the next slide was the title. It was JAWS. And if you liked the movie, the brilliance of Spielberg was like he never showed the shark until later on. It was always like the premise was the audience could imagine it much worse than they could come up with anything on film, so just make it hidden. So, that was a lot in the marketing.

Austen Collins: But then it turns out, most of our investors are Dockers investors. And so, when I was pitching them, I got insecure I took out that first slide. So, I was waiting in the lobby and I thought, "Oh, maybe this is too much." And so, I stripped that slide out.

Adam Elmore: Yeah. Slamming that poor coat.

Austen Collins: So, all that cool stuff never made out there. But that was the beginning of this Serverless buzzword, so a lot of legal threats and realized, "Shoot. We have to change this. What are we going to change it to?" And again, Serverless was not a clear thing. And I had to do the typical stuff of trying to get a good domain and I get all these name spaces. These days, you have to get the Twitter handled and domain, just like the meetup domain. There's just the MPM module. There's so much stuff you have to do.

Austen Collins: And to go through all that stuff, especially get the domain, took me months. And so, I was working on that in the background for a long time. And then finally, by the end of 2015, we transitioned to Serverless framework. And I think the serverless buzzword is a little bit better accepted at that time, but it was a ton of hate that comes with just the innovation process. You're coming out something new. Incumbents are a little threatened, perhaps. And so, but yeah, that was the story.

Austen Collins: And I think beginning of 2016 was when AWS was like, "Oh, there's a major category here. And this could be how we differentiate AWS over the next 10 years." These companies, they make these large bets and this transition to these serviceful architectures and using all these different things is just such a natural evolution for the Cloud. And I think AWS caught onto that. And saw what a lot of what we were doing on the storytelling category creation side. And I think they just redirected a lot of marketing budget in 2016. They're like, "Okay, let's move these millions of dollars towards this word. Let's aim it at this word."

Austen Collins: And it seemed like 2016, late 2016, we were just in full fledged, new category. So, didn't come up with the name. It was out there. Just maybe had something to do with popularizing it. And making the pattern accessible to developers and just making it fun and exciting.

Adam Elmore: Yeah. And so it worked, I think I saw the Datadog state of Serverless report that 90%, this blows my mind, 90% of cloud formation usage, teams leveraging Cloud formation do so with the Serverless framework. That's insane. That's some validation that you did something right.

Austen Collins: Totally. A lot of credit goes to the community of that project of Serverless framework. It goes to AWS for innovating the underlying Lambda. And it goes to timing and luck is a huge part of it, and certainly a huge part of it. And then maybe, I came in with this weird background of design, storytelling, and engineering. And so, it's got this perfect storm and Serverless framework has been on a tier since 2015, since that launch date, just like NPM download growth.

Austen Collins: Everything has been, even 2020 through the pandemic, was I think, it was our best year in terms of growth. There's just like just that clear curve there. And so, it has been, I can't explain it even at, at this point today. It's just, it's-

Adam Elmore: Well, I want to dig into that a little bit. So, six years you had the hot start where you put it out on the internet and it resonated today. And we'll get to some of the latest stuff you guys are doing. Clearly, just a major player in all of serverless, lowercase S. But has there ever been patches? Because you built a company that initially was built on an open source framework, have you ever had tough patches as a founder in terms of, I don't know, monetization or any? I've done the founder thing and raised a bunch of money and we never figured out how to make money. Has that ever been a struggle for Serverless? Has it all just been a pretty smooth sailing path?

Austen Collins: Startups are never smooth sailing, never. That's one of the reasons why moving back to San Francisco, I want to be back in that city after temporarily relocating through the pandemic because startups are brutal and you just need shoulders to cry on. And San Francisco is good for that. There's a lot of people who are just going through it. It's different, what you see on Twitter. Unfortunately, a lot of it isn't so truthful as what you get when you're in an entrepreneurial community where people are, you hear the back channel. You do those coffee meetings and whatnot.

Austen Collins: And so, anyway, but startups are, are tough. It's been a very interesting journey, especially with AWS simultaneously making such a huge bet on Serverless. Right from the beginning to be caught up in all this. Maybe a memoir one day would be fun to write because there's so many stories I probably can't get into on this podcast, but it was a very much just like this storm of something's happening. No one knows what it is and everybody is starting to get a little crazy. But yeah, the open source go to market strategies are very interesting. They're increasingly popular, but the jury's still out. Is this the way to do it? Is this a silver bullet solution for, for enterprise adoption, for winning over mind share?

Austen Collins: And so I think for us, it took us a little while to figure out where the right places are to monetize, in addition to just waiting for the category to grow the overall movement. Because we are we're serverless-tic. We are a purest bet on the serverless category. It's awkward because we're named after the category, which is like it's one of those things. Again, in the beginning it was not clear that this was going to be such a big category. And now, it's as if we make cars and our company is called Car. It's very interesting.

Austen Collins: But yeah, it took us a little while to figure out where the monetization points are in this. And it's specifically, like as a lot of startups, developer tools, Cloud startups learn there's the product you offer to developers and then, but the decision makers, the buyers are typically looking for something a bit different. And it's so easy to get caught up with just let's keep building stuff for all the developers and give them the ultimate power. But for us, it's like developers use Serverless framework every day.

Austen Collins: There's been millions of apps launched with that. Tons of startups, all building in Serverless framework, but people call us. The team leads or decision makers call us because they have scaling challenges. They're trying to grow their serverless team, specifically. And it's still a new architectural pattern that's evolving all the time that's very different. The Cloud services, the configuration is a very different from code. And so, how best practices scaling a team, getting their team organized typically that's when people will reach out to us. And then on the monitoring side and debugging and troubleshooting.

Austen Collins: And it's like, yeah, those, yeah, it seemed pretty clear to draw the line right there because that's when the team leads, the managers, the VPs, CTOs would be like, "Okay, we need some help here." So, we're doubling down in that direction. We have originally with Serverless framework dashboard, which is a SaaS product that accompanies a Serverless framework and it tries to focus on those areas. And we're revamping the whole dashboard with Serverless console, which we actually just announced last week.

Adam Elmore: I saw that.

Austen Collins: Yeah, yeah, yeah. Figuring that stuff out takes time and with a lot of the developer folks companies, it's hard to balance. How much are we doing for developers and whatnot and how much are we doing on the monetization side. And we had a lot of investors from Docker who are very community-focused. And so, for the first few years of the company, we just focused on community because our investors from Docker were just like, "Just grow that community. Yeah, turn it into something that was so huge."

Austen Collins: And I've got mixed feelings about that advice these days going through it, but yeah, that's our story. And I say we definitely finally have figured it out and cracked them up there. But yeah, it's a whole journey with a lot of ambiguity and trailblazing to get to that point.

Adam Elmore: And I want to talk about those new things. I think that's a good segue. So, you just recently launched two big products in Serverless Cloud and Serverless Ops Console, so I'd to talk about each of those. I don't know if maybe it makes sense to talk about the console given that we just were talking about the console. Could you explain to everybody on the call, what is the Serverless Ops Console and what need does it fill?

Austen Collins: I might back up just a little bit here for Serverless Ops, for the Serverless Console and Serverless Cloud, because things are changing in Serverless land. They're always changing. They're always evolving. We've been at this category for six years and right now, we're caught between people need, especially leads need great tools to debug and scale their teams and scale their projects and whatnot. And then also, our perspective of the company is serverless still needs to be easier, which led into the Serverless Cloud. And it's interesting, so we're straddling these two things.

Austen Collins: So, we want to move the category forward, the movement forward and make this a lot easier. That's where Serverless Cloud comes from and then Serverless Console for like, "Okay. There's a lot of problems that are always going to be here with this type of architecture. Let's solve those better for teams." Our perspective right now, yeah, we launched a couple products. Our strategy is Serverless framework, which we have an announcement coming for that B3 next week, so that's been in the works for a while, a lot of improvements.

Austen Collins: And just, yeah, throughout the company, I think we've got one of the best teams we've ever had across all these different product areas. And so, Serverless framework B3 is coming out, so, continuing to invest in Serverless framework. And then Serverless Cloud is a bit of like where could this go next? And then Serverless console, again for debugging and teams and whatnot. So, right now, focused across those three strategies. And I think Serverless has come pretty far. And we're not sure, we're just placing a lot of bets, but we definitely want the movement to grow. And in terms of the fundamental architecture, it still feels like I'm having a lot of the same conversations I had in 2015.

Austen Collins: And that is we're unbundling these monolithic apps and outsourcing a lot of the functionality. There used to be in code to these auto scaling pay for use Cloud services to reduce total cost of ownership. And that's amazing. And these new Cloud services are coming out and these represent perhaps the greatest building blocks of developers of all time, but there's a lot of services these days. I think serviceful when I first heard that term by Patrick at the original Serverless comp, I thought this is, yeah, I love this name, too, for the architecture and whatnot. I didn't know serviceful was going to have, just so many services in the future, a few years down the road and whatnot.

Austen Collins: And going back to what are the challenges of Serverless to solve today? I think right now we have these distributed systems and people are still having a hard time building them, developing them, teams collaborating on them, monitoring them. And we have more services than we perhaps imagine like a classic example is RDS data proxy. I'm talking to customers and users every single day. And just once we get into if they're new, once you start saying, "And then you bring in data proxy if you want to connect to RDS or something like that," you just see their eyes glaze over and go, "What? How many things do we have to bring in here?"

Austen Collins: And so, it's getting interesting. There's still a lot of challenges to solve there. So again, we're trying to improve that. The way it is today while also innovating in Serverless Cloud. My customer put it really well recently. They said like, "I used to spend two months coding my application. Now, I spend two months configuring them." And we thought, "Yeah, yeah, that's right." How do we get that down to a week to configure this stuff and deploy it? And I think there's still a lot of work in the movement to figure out how to do that. And I just hope to God the answer isn't velocity templates.

Adam Elmore: As someone that writes a lot of velocity templates, I can agree with that statement 100%.

Austen Collins: It's interesting time. I think Serverless come a long way. We're ready for, more innovation here. So, anyway, Serverless Console is here to help teams with Serverless as we know it today. That is, it's solving two problems. It's a bit of what AWS is always doing and just getting your AWS accounts organized, your stage is organized. We have a concepts of apps and stages. And the first call we get is from the team leads saying, "How do I set up my team and access and authorization and whatnot. And make sure all the developers are doing this safely."

Austen Collins: And so Console is there to help with that, as well as on the monitoring and debugging side. So, it's a bit of a AWS, SSO overlap with Datadog for the Serverless era. And we really want to position it as, like it's time for a new operations console for Cloud. If you're building applications on AWS today, we're just focused on net new stuff right now. If you're building applications on database today, you're going to be using a lot of Serverless. Infrastructure and containers that run in a Serverless S fashion, yeah, we want to come out with the Ops Console just for this era, for that era of applications.

Austen Collins: And so this is the first few places we're dipping in to do that. It's Serverless Console today, but Serverless and Cloud continue to merge as one thing. Serverless is just like the evolution of Cloud. So yeah, we're hoping that console is what these new teams, these startups use. It's all they need to really get going with the high velocity.

Adam Elmore: Yeah. I've had a lot of guests on the show. We've talked about the shortcomings of the AWS Console and how it's really just about they have to serve all these masters. They have to have a solution that works for every use case. And what you're building is the solution for Serverless use cases. There's a ton of value there.

Austen Collins: Yeah. And we're entering a world with more Cloud than ever. More Cloud accounts, AWS well-architected recommends separation of accounts per stage and workload now. So, you've got all these Cloud accounts, you've got more Cloud services that you're using. It's time to rethink our Ops Console and team collaboration across all these things, because there's probably only going to be more in the near future.

Adam Elmore: And then Serverless Cloud, is it aimed more at people that have less knowledge coming in? Is it meant to abstract a way infrastructure generally or is it, could you elaborate on maybe who it's targeted at?

Austen Collins: Yeah. It goes back to the challenges we see with Serverless, yeah, still. And that is, "Okay. There's a lot of services." And I go back and forth with the Serverless community on this. The notion that if we're going to increase the Cloud services that you use and have just low-level granularity of all these services, you need to use the notion that we're going to bring everybody into this world and make them a serviceful expert. And have them practice, know all these services and the ins and outs of them and get security right across them and just collaboration and the life cycle of applications, I think is a long shot.

Austen Collins: I think that that might be unhealthy for the movement. And we're almost creating this new ivory tower of serviceful experts, which I just, I hear the AWS community talk about this and say all the right things. But every single day, I talk with teams, who are just getting ramped up on this. And there's such a big gap still between what the AWS community is talking about in terms of best practices and "you should know all these things," to what these startups, these new innovation efforts in large enterprises are dealing with.

Austen Collins: And so, I think we need to, yeah, we got to solve this to move the movement forward. And so, what does that look ? Where is this going next? It's tough. That's something we all have to figure out. But our take on Cloud is your number one higher levels of abstraction. At the end of the day, to make a lot of these services go away, it's just time for new levels of abstraction. And so, Cloud does introduce those and our hypothesis there is the future of cloud is focused more on outcomes and used cases, not on infrastructure. For instant business solutions out of the box and that's where a lot of this stuff is going.

Austen Collins: And so, Serverless Cloud introduces some new higher levels of abstraction for common Serverless patterns. We don't have to know all the services at the end of the day. And it is trying to remove a lot of that configuration complexity and automate more of it. So, high levels of abstraction and then better developer experience around it so that the classic challenge of serverless architecture is that you lose the pace of productivity that you had when you just spun up this code based on your local machine.

Austen Collins: Code, where we transition from code to infrastructure services, you just lose a lot of stuff. Code as much better DX than infrastructure. You lose speed, time to action, forking, branching, configuration, developer autonomy. Collaborating on shared Cloud infrastructure for a large team is hugely painful. There's a lot of issues. Testing, there's just so many challenges there, so we're trying to bring that DX back to infrastructure. And so, the way that we're doing that is solve the trifecta of developer bliss, which is fast time to deploy, fast time to interact with the thing and fast time to get the feedback, which is the locks.

Austen Collins: So, as Serverless Cloud high levels of abstraction that deploy in 500 milliseconds and a log stream back instantly to the Console. And so, but it's all on the Cloud. You're working on the Cloud and I know there's that the classic is local the path forward for Serverless or is development on the Cloud? I've been on the Cloud development train since 2015. We've done a lot of work with Serverless framework to do offline emulation and all that stuff. But it always breaks down for larger teams collaborating and as soon as you add more services, it really, it breaks down.

Austen Collins: It's easy to get it up and running with smaller projects and smaller teams, but scale is really hard for offline emulation. So, we got to bring development to the Cloud, but we can't do it unless we recreate the speed of local. So, that's what Serverless Cloud is trying to do very well and the team has nailed it. The engineering work that's gone into upload caching of your code, to get those deployments feeds down and just, there's just so much great work that's gone into it. And so, they've really nailed it. So, we're bringing that pace of development back to the infrastructure.

Austen Collins: And then getting into some interesting stuff around the forkable Cloud, which is, I think recreating a lot of the stuff we used to have in code branching and just cloning, the forking of the code base, and iterating on it. With Cloud, it's the first time where you actually clone the infrastructure, the code, and even your data in 5 seconds. And, Jeremy Daly, who's leading the effort over there and the team have come up with the most incredible vendor management account, vending system in Cloud, where literally every time you fork your application, it spins up a new AWS account. Deploys, the infrastructure and the code, and even copies over data optionally in five seconds, which is insane.

Adam Elmore: It's insane. Yeah.

Austen Collins: And we're not talking a lot about this yet, but the way I describe it, I say like, "Hey, you guys, you're treating AWS accounts as the new container and you just built Kubernetes or something." Yeah, but it's very cool. Every app stage or we call an instance, you could easily fork it and spin it up in a new account. Total isolation. There's no shared resources. You reduce a lot of security constraints and you could do it fast. And this solves a lot of problems.

Austen Collins: What does this whole forkable Cloud solve? It's first-off developer autonomy. If you're a developer and you're trying to collaborate on a large set of Cloud infrastructure, and you're working on a shared development environment, you could do individual developer sandboxes, but then how do you recreate all that infrastructure in your sandbox? What does that look ? Hugely painful. Or if you collaborate on a shared environment, everybody's colliding. They're breaking stuff, they're blocking everybody else. It's really frustrating.

Austen Collins: Again, this is the stuff we lose when we go to infrastructure from code. So, the fact that a developer could just clone any stage, they could clone off fraud, which might be a little crazy. But if they have a shared like test stage, any developer could just clone that and work with a lot of autonomy, innovate, clone the whole application, all the infrastructure instantly. So, it brings that autonomy back to the developer workflow.

Austen Collins: And then you have testing, which is the next one. Test environment, someone breaks, it blocks everybody and, hugely painful. Cloud, actually, you could spin up when you run your integration test. You could spin up just clone the application. Run your test against it and destroy that clone, again, in five seconds. So, that's in its own is a revolution. And then, there's preview apps that you could create. Spinoff really quickly to share feedback and whatnot. And then there's a lot of social community based features where anyone now can actually clone a public Cloud app.

Austen Collins: And it does a lot of stuff where you don't need to get, so it will clone the code that will deploy the whole application. And set up a session where it's streaming logs, watching your changes deploying instantly. And all this stuff you get with this forkable Cloud concept. And so, that's I think it's the high level abstractions, developer. The pace of development is back with real live Cloud apps and then this forkable Cloud concept. Those three things I think are killer. Definitely feels next generation Serverless development.

Adam Elmore: It was a huge step forward. And I think it's interesting the way these episodes of AWS FM have timed out. I had a whole bunch of guests early on saying, "We should be doing all of our dev against the Cloud." That's what they've been saying. That's what you've been saying for a long time. But Serverless Cloud wasn't necessarily launched yet or it was still so early, they hadn't played with it.

Adam Elmore: And then I had Brian LeRoux on my last episode, who's been the other side of that debate. It's a smaller group that's saying we should emulate, we should do things locally. And one of the questions I asked Brian was from Ben Kehoe, who asked, "If we could test in the Cloud, if we could do our local dev in the Cloud just as fast as you can do locally, is there a need for that local emulation?" And that's here now. You, guys, you've answered that question. And it's just so new that this conversation I've been having on the show, it already had an answer. We hadn't gotten there yet, so I commend you for moving the whole movement forward so quickly.

Austen Collins: Well, it's hard to replace local, to get that speed back, to be frank. It's one reason why we had to go down the custom abstraction path to do the whole forking and all this stuff. It's tricky. We really had to come out with our own Cloud services as a result. So, it's easy to say, "Develop on the Cloud." It's hard to do that for teams because it's too darn and slow.

Austen Collins: Most people try Serverless and they're like, "I'm going to wait a minute for Cloud formation to deploy every time I do a code change? Are you nuts? I'm out of here. Don't tell me about the Serverless stuff ever again." And it's unfortunate. So, bringing that speed back has been hard, but I think it's the only direction to go. And look I've seen teams spend more effort trying to sustain their local emulation and recreate AWS locally to the point where all these, the time savings, productivity savings of adopting Serverless in the first place are just out the window because they're all trying to maintain this massive local emulation.

Adam Elmore: Yeah. And it's a moving target, yeah, so it's a constant battle. Yeah, I think it's really exciting stuff. I love seeing the pace of innovation that you all are pushing the envelope so quickly at Serverless, Inc. It's really exciting for the space. I think AWS is doing things that are really exciting in terms of supporting Serverless. And I think of the step function stuff that's recently out. Are there things, in AWS land, I know you guys, you work with other Cloud providers, but in terms of AWS, are there things that you'd to see from AWS moving forward that would make you feel like enable people like you on the third party side, I don't know. Things in Serverless more generally or just where you think the future is headed from the Cloud provider standpoint?

Austen Collins: I think AWS is an infrastructure is a service company and I hope that they keep focusing on that. Personally, I thought there would be more APIs for business outcomes at this point. I know they've come out with a lot of managed services for still lower level concerns, but the API for everything, the API economy, I'm a huge, huge fan of that. That's very much in the same vein of Serverless. I hope AWS focuses more on that and puts an API in front of every business problem, life problem that people have.

Austen Collins: Because again, I'm a huge believer that these are the greatest building blocks that will usher in this next era golden age of software development. So, just getting more of those things that developers cannot code and maintain themselves and just bring into their applications to innovate better. Would still love to see more of that rather than a lot of low level services for things. And so, yeah, I was hoping that we'd see more of that at this point. It is also getting into dev tools there.

Austen Collins: But when we got started with Serverless framework, AWS's developer tooling ecosystem was nothing. And now, in typical AWS fashion, they said, "Well, we want to do something this." So, then they threw 10 solutions at it. And so now, they have, there's CDK, you've got Sam, you've got just basic Cloud formation improvements there. They amplify there's so much here. I think AWS always has a hard time building opinionated experiences.

Austen Collins: And a lot of that is, I think it might be a cultural challenge where they are an infrastructure as a service company and their model, the repeatable scalable model is we're building services for everything and we're going to do everything that customers ask for. And so to then, to bring in like an Apple mentality where you just want to have these singular beautiful experiences on top of that culture, I think is really hard for them. So, now, we've got 20 different developer tools or experiences or something like that, which comes with its own drawbacks of just saturating the industry, which is too many options.

Austen Collins: Yeah, those are the two things that mostly, just more infrastructure as a service. I think Cloud infrastructure automation has aways to go still. I think CDK is interesting. I'm biased because I made the Serverless framework, of course. I just...

Adam Elmore: Yeah, sure. It'd be hard not to be. Yeah.

Austen Collins: ... get that out there. But in terms of moving this stuff forward, I think, CDK is interesting. I'm just looking for more though, these days than just, when we embraced Cloud formation with Serverlesss framework version 1, I still feel like it was the wrong move, actually. And in Version 0, when we launched Serverless framework Cloud formation, it wasn't in there.

Austen Collins: And the theory was, the hypothesis was we're going to embrace Cloud formation to improve developer experience and what people could do with Serverless framework. Because if Cloud formation handles infrastructure automation, then we'll be free to do all this other cool stuff with the framework, right?

Adam Elmore: Yeah.

Austen Collins: It turns out that Cloud formation is at the heart of developer experience, so it's right. Depending on your perspective, it's good or it's very, very bad. So, I think on the infrastructure side, I don't know, this might be a tooling company. We've tried to experiment here a lot, but I love, I just feel like there's a lot more that we could do here that is more automation out of the box. It's more focused on the app level, on specific outcomes.

Austen Collins: Again, I'm a big on this like the future Cloud is outcomes. It's not infrastructure. That's a big theme of ours, the company. So yeah, I don't know what that looks, but just looking for something that really pushes the boundaries there.

Adam Elmore: Well, I want to dive a little bit more into things about you and less just about Serverless. So, I think before you started Serverless, before you started JAWS, you were a AWS consultant?

Austen Collins: Mm-hmm.

Adam Elmore: Is that accurate? So, did you spend time working with AWS customers, building out solutions for them? Could you speak a little bit to that phase of your career?

Austen Collins: Elastic Beanstalk was my tool of choice. And so, I've built out a lot of projects for customers on that. And this is when I moved up to Bay Area and that was, my primary focus was startups. And so, I did consulting work on the side just to fund startups. And so, it was just a lot of that complexity. Elastic Beanstalk, I think, was supposed to be the Heroku flavor. I've always been a big Heroku fan as well. That's some of the original serverless DNA. James Lindenbaum, the whole team over there were just like so interesting. They had so many progressive thoughts when they were doing all that.

Austen Collins: But it was a lot of, yeah. Elastic Beanstalk and wasn't a big fan. Again, I think that it was too much of a distraction. Ben Kehoe likes to say, "Serverless is all about focus." I'm a huge believer in that. Absolutely. How do you get that time back to focus on those hard problems of product market fit.

Adam Elmore: So, I'm an AWS consultant. And I just, I had to ask, because I wondered, is the trajectory of my career, is my next step to build some hugely successful framework? I'm hoping that's the case. We'll see, but it doesn't sound like the consulting side was a major part of your story. It was a thing you had to do just to pay the bills.

Austen Collins: Definitely. I'm entrepreneur first. I've got a list of startups that's a thousand pages long. This is the greatest time, really.

Adam Elmore: Oh, it really is, yeah.

Austen Collins: It's just like to leverage any. This is one thing that still gets me out of bed in the morning to do this serverless dev tools company is every day I talk to these small teams, who are doing incredible amounts of work. I just cannot believe the amount of productivity that they're reaching. And the valuations that their companies are hitting all through this with really small engineering teams. It's wonderful.

Austen Collins: So, being close to AWS and the power and the leverage that you're getting is super important here. But yeah, the consulting path wasn't an end goal. It's a fantastic, I mean, the consultants right now are just making a killing. In AWS land, if you're truly a great expert there, it's just a wonderful lifestyle.

Adam Elmore: Yeah. The building products is cool, too. I've definitely had that bug in the past, I think. Yeah, there's never been a time this. And I think I was on a walk this morning with a friend and talking through, I just can't imagine for me ever wanting to work for somebody else again, because there's just so many things I want to do and things that I would want to build. It's hard for me to imagine hiring people because I would feel bad for them.

Adam Elmore: I wouldn't want them, I don't know this is going on a tangent, but I can't imagine their motivation to work for me. I just feel like there's so much to build. So many things I'd like to go out and do. That's entrepreneurial. Am I entrepreneurial is what I'm getting at. Does that mean that I'm entrepreneur? Okay, yeah. That's what that means.

Austen Collins: You just got out of a startup, you went into independent consulting. You like to preserve your autonomy and your ability to work on the things that you want to work on and be your own boss. And I think that's increasingly going to be the future, more and more of these. I don't know if they're independent contractors or what, and you don't need it. I just don't see that the large team of folks who could bring in these independent contractors, someone like an expert you, in to make a big impact, incredible impact in a short amount of time using Serverless technologies and whatnot. And there's just going to be more of that.

Adam Elmore: Yeah. Serverless makes it, it's really exciting to leverage a tiny team or a single individual can have with technology today. What about, do you have any role models, anybody growing up in technology or early in your career, any mentors?

Austen Collins: Yeah. Just so many in-depth tools, like there's some masters out there. People that I think are masters of the craft. And I was very much inspired by these people. TJ Holowaychuk in Node land, if you know him. He also built serverless tools as well. But if you look at a person like that, the way they approach getting their dev tool to market. The simplicity, the focus, the attention to detail, the docs, the marketing around it, there's just some people out there who are, I think, everything as a craft in building out dev tools. Especially, there are clear masters.

Austen Collins: But I think, the founder of Vercel does a great job. He's one of those folks, who's just like, "Okay, this is a master at work." And anyone who's just mastered anything in any domain or something like that I love. All those people inspire me. I'm very craft focus. Maybe that's the Hollywood background or something, because everyone always talked about filmmaking and whatnot. Is this craft? And there's this just beautiful culture around it and sharing that information and it's all about mastering that. So, those people inspire me a lot.

Austen Collins: Yeah. Something that's my personal journey going from just engineer to startup founder and CEO here, one thing that's inspired me a lot actually is getting into the different departments and functions of scaling out a company. I talk to-

Adam Elmore: It's a whole another job.

Austen Collins: It's a whole another job, but it's part of the same journey and overall product and company at the end of the day. So, I talk to a lot of engineers who are... yeah, they're always talking about growth, which is important. And they say, "I need to grow, so I'm going to adopt this new language or this new thing or something." And I thought a lot about that because I'm quick to pick up whatever the new thing is. But it's been interesting when you go on the journey to build a company and found something, you have to get out of engineering and get into these other disciplines.

Austen Collins: And so, the people I've met who have helped me understand product marketing sales, all that, has been really inspiring for me. And looking back, I feel like I didn't know a lot about engineering until I got into formal product management and product development. And I didn't know a lot about product until I got into marketing. And then I didn't know a lot about any of these things until I got into sales. Because when you're asking, when you're interacting with customers and you're trying to really sell your solution, that's when everything comes to a forefront. And that's where all the most, I think a lot of the most important learnings are. And I didn't know a lot about any of these things until I got into the management and the CEO position and whatnot.

Austen Collins: And so, the people who have helped me, there's a lot of advisors. Being around the Bay Area has been great because you could reach out to these folks, get coffee with anyone. So, there's a lot of people who've helped me on that journey, which has been tough, to learn all that stuff in the context of a startup and pressure and capital and whatnot. But looking back, I just, I had no regrets. I just feel like that was the best way for me to grow a lot. And it's almost to grow as an engineer, the best thing for me was actually to step out of my lane and get into these other things.

Adam Elmore: The other concerns, yeah.

Austen Collins: Because otherwise, we're back to just discussing the latest state management pattern and react or something.

Adam Elmore: Sure.

Austen Collins: And today, there's just so much churn. There's so much stuff out there and it doesn't feel like anyone is really keeping score in terms of what the new thing is and whatnot. And yeah, I've adopted a lot of things and looking back, I'm like, "I don't know if I made a lot of progress" or if it was worth doing all that stuff. And there's also that school of thought where it's like, "It's not what you got, it's what you do with what you got."

Austen Collins: It's all about the practitioner. And yeah, because a lot of practitioners can take these old technologies and build incredible innovation in companies and whatnot. So, TLDR is just stepping out of my lane. Getting into these other disciplines has helped me, I think, become a better engineer and really realize what matters. What matters most in the innovation process and the people who helped me along the way there were huge. So, that, and the masters of the craft in-depth tools and anywhere.

Adam Elmore: Okay. Last question for you. Do you have any hot takes? Hopefully, technology, but any hot take I'll take.

Austen Collins: Yeah. So, we have to enrage Twitter here. Is that the same?

Adam Elmore: Yeah, no, that's the goal.

Austen Collins: Is that actually correct?

Adam Elmore: Yeah, yeah.

Austen Collins: Not so much. I'm optimist, I'm an optimist at the end of the day.

Adam Elmore: Yeah. I think you have to be an entrepreneur, especially a solo founder. I think if you weren't an optimist, this would not have done well.

Austen Collins: You're in trouble. Totally, yeah. And I think that there's a lot of more than ever, maybe it's always been this way, the social media services, but it's people are throwing shade left and right on all the things. And it's so interesting. With even CDK coming out and what people are saying about it and then Web-3, that trend and you see. The most fascinating thing to me is people who've been through a part of something that is an innovation, great innovation and then they're so quick to say, "But this next thing sucks."

Austen Collins: They went on this journey and they're part of something and hopefully, they saw the innovation process, which is messy and the early versions of things are totally screwed up and nobody knows what they're doing and it's just like an anarchy. And then it turned. And then, eventually you convert that ambiguity and that chaos into order through that process and it becomes something. So, a lot of the new stuff, yeah, even if I don't understand it that much, I'm open to it. I really want to figure out what is it that's happening.

Adam Elmore: That does feel like a hot take for what it's worth on Twitter. To be okay with new things that are presenting to.

Austen Collins: To be okay with new things? Yeah.

Adam Elmore: Yeah, that sounds like a hot take.

Austen Collins: Yeah. One other potentially hot take is a philosophy on things on dev tools and so much more. From the beginning, a lot of times I can't explain why things get popular, to be frank. I just don't know. I don't know. I can't explain Justin Bieber. I just can't. There's a lot of stuff that just happens and it becomes popular.

Austen Collins: And Serverless framework to some extent. I'm just so grateful for the journey and just going through this experience. First off, if you're a maker, there's nothing better than seeing people use your thing and having a great outcome. It's just like that is your proud parent moment. It's just, I live for those moments. It's incredible. I really appreciate just the people who've tried it out and the community around it.

Austen Collins: But I can't explain. Where is this? Some days I look at the download charts and I think, how is this still happening? I can't. It's so interesting. And when it comes to dev tools and engineering patterns, best practices, new Cloud innovations and all that stuff. I can't tell if ideas come to the forefront because they're the most logical or the most fashionable. Yeah, that is something I battle with internally all the time, because we're seeing so much these days. And it's tremendous innovation and I know it's a lot.

Austen Collins: And it's like, "How are these things coming to the forefront? What's driving these? Are these truly better? Who's keeping score? Who's doing the measurements? And is this stuff, is there something that's fashionable about it? And in engineering culture, it's hard to say this, I think because we're obviously also rational and we never make emotional decisions. We think through everything.

Austen Collins: But then you look at what engineers do on Twitter and there's just emotion behind everything at the end of the day and whatnot. And I'm jaded, because I grew up in Hollywood and that's the fastest place to lose your innocence. It's like if you grew up in Hollywood, politics or media, any of these industries and you see what really goes on behind the curtain and just the manipulation at that, it's an engineering discipline on how people craft these experiences, these stories, these narratives and whatnot.

Austen Collins: It will change your perception on a lot of things. But yeah, maybe that's, that's typically a hot take in engineering culture because we typically like to think, oh, we're, we're the most rational creatures around and we think through everything and whatnot, but sometimes things are just cool and we like them for those.

Adam Elmore: Yeah. We get swept up in the stories. Yeah, absolutely. Well, Austen, it's been so good to have you on the show. I learned a lot today. Every time I do one of these shows, I learn a lot, which I think underrated way to learn things in 2021 is start a podcast. I don't know, maybe everyone knows that and I just learned it, but-

Austen Collins: I've got a growth hack, too. This is just like a random piece of advice. For people who are on the startup journey, that is just in the same vein, one of the best things I've done as with building a company is doing the executive search. And you go in...

Adam Elmore: Nice.

Austen Collins: ... and if you have the venture capital, you have the capital, you go and you retain an executive search firm. And it's going to be pricey, but that executive search firm and you're trying to find C-level VP level position, but that executive search firm is going to put you in front of these amazing people, amazing people. And just to sit down and have conversations with them over a couple of hours and the data. The things that they've seen. Again, it can be hard because a lot of stuff that you read on the media and Twitter, yeah, this might shock people, but not everything on the internet is real.

Austen Collins: And so, to be able to get in a position where you have these long format conversations with people who've done it. Who've been through it, who've seen the internal, seen what it looks from the inside, they have data, they've got metrics on whatnot. And they speak from a position that is hopefully is informed by that data. It's the best thing. So, whether you start a podcast or if you're just doing a startup, consider stretching out that executive search, depending and just doing that on a recurring basis.

Adam Elmore: Just tuition. You just get to, yeah, learn from the best.

Austen Collins: Totally.

Adam Elmore: Yeah. So, a couple things join the Discord. Austen, I'd love to see you on the Discord. We've got a Discord. There's a bunch of the former guests on the show, a whole bunch of listeners and we just interact and talk about how we can make the show better. I'd love to see everyone join the Discord. If you listen to the show on a podcast platform, I'd also love a rating and a review. That's great feedback for me. I like to know that the show is resonating with people. It helps me keep doing it and it makes me feel good. So, I'd love to see those reviews. Yeah, that's my podcast plug.

Adam Elmore: And Austen again. Thank you so much for joining. It's been a great time.

Austen Collins: Likewise. This was a ton of fun. And you're doing a great job, by the way, so keep it up.

Adam Elmore: Thank you.

Austen Collins: I'm already a listener.

Adam Elmore: Oh, that's fantastic. The number of people that have told me that they're a listener, it blows my mind. I don't know why people want to hear me talk. And it's hard to know from your own perspective why people want to hear you talk because you're you and you hear yourself talk all the time.

Austen Collins: It's very casual. It's very freeform. I personally like that a lot. And I think you just have a great energy.

Adam Elmore: Yeah. It's not me talking. It's the people I've had on the show are talking and they're fantastic. So, yeah, in that sense, it's a great show. Listen to the show.