Filling the Cloud Talent Gap

Transcript

Jared Reimer: The demand for experienced competent cloud talent continues to rise much faster than the supply does.

Rahul Subramaniam: The cloud environment is changing so rapidly, that keeping up, in itself, is a mammoth challenge.

David Brassor: Where is the value-add of leveraging a third party service provider? And where should I be leveraging my own teams internally, moving forward?

Hilary Doyle: So that’s a great answer, but I also recognize you’re a consultant and that is a very consultancy answer that you’ve just given us.

David Brassor: Have you been there? Have you done it? Have you got the t-shirt that says you’ve done it? The answer, in many cases, is “No.”

Hilary Doyle: This is AWS Insiders, an original podcast from CloudFix, bringing you what you need to know about AWS through the people and the companies that know it best. CloudFix is the nonstop, automated way to find and fix AWS-recommended savings opportunities. It literally never stops. Neither do I. I’m Hilary Doyle. I’m the co-founder of Wealthy Works Daily.

Rahul Subramaniam: I am Rahul Subramaniam. I’m the founder at CloudFix and a FinOps practitioner and enthusiast.

Hilary Doyle: Rahul, they tried to hire AI co-hosts to replace us this season, but no algorithm can nail your comic timing. And frankly, none of them would do my puns. So we’re back.

Rahul Subramaniam: Yes, we are. Rested and ready to break down all the latest at AWS.

Hilary Doyle: I don’t know about rested, but listen, the cloud moves fast and we have got to keep up. So season two, we’ve got it on lock. Here we go.

There’s been a lot of talk about staffing shortages over the last few years. They’re manifesting in all kinds of ways. I mean, maybe you’re waiting longer at the grocery store checkout, or your favorite restaurant may be closed a few nights a week because they can’t staff the front of house. In healthcare, obviously, this issue has been acute among frontline workers, pharmacists, GPs. No one is immune. Not even, as it turns out, the cloud.

The International Data Corporation, sounds very serious, they predict a global shortage of 4 million full-time developers by 2025. And so for companies moving to, or already in the cloud, there is a very real challenge, some call it a crisis, of finding, training, and then hanging onto your cloud talent.

I mean, didn’t we think that moving to the cloud was the hard part? And now it turns out it’s also hard to stay there.

Rahul Subramaniam: Absolutely right, Hilary. Making the decision to move to the cloud is literally just the first step. The real challenge is when you start executing that move. Staffing your cloud team with the right resources is probably one of the most critical steps right now, and also, possibly, one of the hardest.

Hilary Doyle: Yeah, they definitely buried the lead in the cloud manual on this one. But anyway, this is an issue that a lot of companies are struggling with. Everyone from small startups, to huge corporations. I’m wondering if it’s fair to say that the most significant challenge is that the cloud industry is moving so quickly, it’s hard for expertise to catch up to it?

Rahul Subramaniam: I think you’re spot on with that one. Change is happening on just so many dimensions. AWS services evolve every week, and then new services are created all the time. And also, at the same time, new patterns and best practices are constantly evolving. That just creates a new type of staffing challenge that, to be honest, even I’m grappling with.

Hilary Doyle: Oh my, God. Have we reached your Waterloo? Because I have been waiting for this moment. In fairness, Rahul, this is a challenge that literally no one has completely figured out, and so I’m glad that you are moderately grappling with it. But just like all business challenges since the dawn of time, it’s going to take some creativity to solve. And so we have gone looking for inspired ideas about cloud staffing and the people who can speak to them. That is all coming up.

But first, Rahul, I’ve never actually heard your cloud computing story. You were a very early adopter. How did you even know what you were looking at

Rahul Subramaniam: Hilary, I’m about to reveal that I’m really old.

Hilary Doyle: I am older than you, so please tread carefully.

Rahul Subramaniam: So I started with AWS back in 2007, and I don’t really think the term cloud was used to describe those handful of API that existed back then. If someone asked you what types of cloud were there, I think the only reasonable answer you would get was the likes of cirrus or nimbus. Am I remembering my clouds right?

Hilary Doyle: Yeah. Or stratus. You were basically an adopter when the cloud was just a misting?

Rahul Subramaniam: Basically, right. Anyway, coming back to your original question. When I first started with AWS, like I said, there were just a few API, and this was for EC2 and S3. There wasn’t even an AWS console. No IM and then no VPCs.

Even then, I thought AWS was a game changer because, to me, it meant no more data centers or managing those crazy ESX servers.

Hilary Doyle: If we could go easy on the letters, we’ve got a lot of people listening who speak in words, R.S., but you know what? While we are on the letters, let’s check the mail of the day, by which I mean the AWS news headlines. How is that for a pivot?

AWS has just announced that Amazon’s SageMaker Inference Recommender, or IR, has deeper integration with Amazon CloudWatch for logs and metrics. And, because of a new API, there’s support for running load tests on existing endpoints. Rahul, you look unusually excited about this IR news. Why is it a big deal?

Rahul Subramaniam: So Inference Recommender has been around for a little while and it really started easing the burden on developers. But then some customers seem to be very perplexed by why the Inference Recommender was making the decisions that it was making. 

Sometimes it would scale up like crazy, sometimes it would scale down like crazy. It was basically a black box. So these latest updates make the recommendations more transparent to the end user, and most importantly, it builds trust in the system.

Hilary Doyle: Okay, another AWS improvement that’s just been announced is all about EBS and AMD processors. EC2, C6A, and M6A instances are going to have much faster EBS volumes. I’m sorry. I mean, faster is better here, right? Can that just be my impressive tech conclusion?

Rahul Subramaniam: That’s absolutely right. But I’ve really been waiting for this one for quite a while. Let’s just take a quick detour and understand the background behind this one and why this is so important. 

Hilary Doyle: That’d be helpful. Thank you. Yes.

Rahul Subramaniam: Okay. So Intel and AMD. I mean, rivals, they’ve been going head-to-head for years, and with each generation of processors, we, the customer, look to see who’s ahead. On top of that, AMD processors have been priced roughly 10% cheaper than the Intel variants, specifically on AWS.

Hilary Doyle: Oh, wow.

Rahul Subramaniam: Now, when Intel and AMD released the fifth generation of AWS instances, or the last generation of chips, AMD was about 10% cheaper, but the Intel chips actually performed better, for most general purpose workloads. So sticking to Intel just made sense. And of course, Intel also happens to be default on the platform. But all that changed with the sixth generation of instances where AMD really started kicking Intel’s backside.

Hilary Doyle: Whew.

Rahul Subramaniam: Just on pure CPU performance. Of course, alongside, you also had these instances being 10% cheaper, right?

Unfortunately, there was this limitation of the AWS Nitro system that caused the IOPS, or the input/output operations, on these attached volumes, which is the storage, which basically caused the throughput to be throttled on the AMD instances.

So even though we have better and cheaper CPUs, customers now needed to make this difficult decision, about whether this IOPS throttling was going to hurt them or not for their workloads.

The vast majority of customers just played it safe and stayed on Intel.

Hilary Doyle: Right.

Rahul Subramaniam: With this announcement, that decision to move from Intel to AMD just became a no-brainer. Almost all applications should work just as well, if not better, on AMD. And it’s going to cost you 10% less. I mean, it’s a welcome simplification and a win for the customers, who are trying to get more efficient.

Hilary Doyle: Okay. Speaking of wins, I know that you are a big World Cup fan, Rahul. Cricket World Cup.

Rahul Subramaniam: You got that right.

Hilary Doyle: Okay. Well, this next story is about that other World Cup, eTF1. AWS was essential in sharing every kick, every weeping player. This is the most watched sporting event in the world.

Rahul Subramaniam: One of the most, Hilary. One of the most. I mean, a single India/Pakistan match ends up having at least a billion viewers.

Hilary Doyle: I just want peace with you, my friend. But seriously, you’ve got folks watching on smart TVs, on their computers, on their phones and their tablets. That means huge connection and authentication peaks, a number of times a day. There is no room here for low latency.

Rahul Subramaniam: Absolutely. I mean, leaving the football versus cricket debate aside-

Hilary Doyle: Oh, never.

Rahul Subramaniam: … And coming to TF1’s challenge. Anyway, way back in 2019, TF1 had to start planning for all of this. Everyone knew when the World Cup was going to be. And this is a story that I find fascinating.

For instance, TF1 one adopted a microservices architecture deployed on Amazon Elastic Kubernetes Service, or EKS. They designed a set of REST and GraphQL API for the front end, and for maximum scalability and event-driven architecture with Apache Kafka. So I wonder why they didn’t use Fargate or Lambda. That would’ve been even neater.

Anyway, they brought in multiple content delivery networks, including Amazon CloudFront, to reliably distribute the video streams to client devices. Thanks to all of that and more, millions of people around the world got to see Kylian Mbappé score a hatrick and, unfortunately, not win the World Cup. I was totally cheering for him.

Hilary Doyle: Well, the world can be a cold and unjust place, Rahul. Those are your AWS headlines.

Never stop recruiting better talent. That is one of the leadership philosophies of Jared Reimer. He’s the company founder, president and CTO of Cascadeo. They’re based out of Seattle and Manila, Philippines.

Cascadeo is a leading cloud IT transformation services provider and longtime AWS premier tier services partner, but they don’t just strive to recruit better talent. They’ve also taken cloud talent development into their own hands. Launching later this year. Cascadeo University.

Welcome to the show, Jared.

Jared Reimer: Thank you very much. I’m glad to be here.

Hilary Doyle: We’re very happy to have you. I mean, first things first, we’ve heard a rumor. Can you explain the connection between hiring cloud talent and fire bubbles on the Mekong River?

Jared Reimer: Yeah. So very, very early on, in the formative days at Cascadeo, we were hiring folks locally in Seattle, some of whom we’d worked with before. And I was traveling a lot back then, before I had kids, and was in northern Thailand on the border with Laos. And I was there to see this event that periodically occurs, where there’s some kind of methane bubbles or something, that spontaneously combust, coming out of the river. And I did not get to see the fireballs, unfortunately, that year. It didn’t seem to happen. Unclear why. Maybe it’s just bad luck or maybe the whole thing’s an urban legend.

But I was there and I was sitting on the edge of the river. And I thought maybe there’s cloud talent elsewhere. And so I put up some ads for talent in India, in the Philippines, and in Thailand. And ended up hiring two of the folks that I met through that process.

Hilary Doyle: So no fire bubbles, but thought bubbles instead?

Jared Reimer: Exactly. Yeah. And ended up hiring two really amazing engineers that way.

Rahul Subramaniam: Wow. So Jared, what got you thinking about extending your talent pool globally?

Jared Reimer: It wasn’t really the cost thing. So if you recall, back then, there was a lot of kind of cost arbitrage with offshoring software development and things like that. But for us, it was really scarcity of talent. And that actually continues to be one of the biggest challenges for us. And that’s true around the world. 

So ignore the inflation and recession and whatever else, the demand for experienced, competent cloud talent continues to rise much faster than the supply does. There are a lot of folks that, maybe they were VMware, IT people, who kind of have tried to rebrand as cloud native or cloud First, but those really aren’t the same thing. And so what we were looking for was raw material, and we found it. So young, smart, ambitious, ready to learn. Maybe we don’t have to break them of bad habits from the virtualization era, where cloning VMs and moving VM images around was the mechanism. 

We really wanted to build our own army of cloud native people. And I’m really proud to say that we’ve done that. And this year we’re launching Cascadeo University, which is our initiative for workforce development.

Hilary Doyle: I have some questions about Cascadeo University, specifically because I think we often talk about things that have already happened as obvious solutions, and then we never talk about how they actually happened. I mean, what you’ve instituted is sprawling and obviously very effective. Can you tell us about the process of building this training system? Give us a sense of the how.

Jared Reimer: Yeah, I mean, I think the key thing for us is great leadership, great mentors, top talent from top computer science programs as raw materials. And then, honestly, AWS has done such a good job with their training and certification. Starting from entry level all the way to the top of their certification hierarchy, it’s really incredibly well done.

And I think that was not an accident. I think they understood, I’m speculating here, from a very early stage that if they were going to win the long game, building a competent labor force that understood how to use the platform and not just host VMs on it was essential.

So we very closely and meticulously adhere to their training and certification process standards. We encourage folks to take additional certifications and specialties. We pay them the whole time. So they’re not expected to do this on nights and weekends. We don’t take that away from their PTO or expect them to cut into their family time with it.

To us getting those certs, getting that experience learning, that’s part of the job. It’s not something you do to get the job. And it never stops.

Like I, myself, keep taking more of these classes. And I just myself retook the Pro Architect exam. It’s expected that if you’re part of the engineering organization, that you’re always upskilling, you’re always progressing. That’s just part of the culture.

And if you’re not, what are you doing? Are you just on vacation the whole time? That’s not really the culture here.

Hilary Doyle: Okay, that’s so interesting to me. I mean, what kind of impact does having to run your own talent development factory in-house have on your bottom line?

Jared Reimer: Well, I mean, it’s a long play. So if the goal were to optimize for profit, you wouldn’t do that, if the goal were to build something sustainable and to do well by doing good and breed loyalty. And in our case, a big part of it is, unlike most of the foreign companies that operate in the Philippines, we’re not there to pull the talent out of the country. We deliberately want to keep the talent in the country. And we do have people in North America, generally people that are client interface, like engagement managers and the senior architects, the leadership, mostly are in continental US, but most of our fulfillment and delivery capability for professional and managed services is done out of the Philippines.

And we’ve found that’s a really great setup. We really like that model, and our clients like it, too, because we can provide resources that are oftentimes much better than what they might find on the local market, if only out of scarcity. There’s just not enough cloud talent out there to meet the remarkable increase in demand.

Rahul Subramaniam: Just taking a step back, what exactly is the skillset that you’re looking for when you talk about cloud talent?

Jared Reimer: So the really large global professional services companies had to scale their workforce to hundreds of thousands of people, potentially. And there’s no way to do that from zero, fast enough.

So what they did, and what most of the folks out there have done, is take experienced IT professionals and try to retrain them to do things that are different than what they’ve done all their career. So if you were a VMware-certified IT operations professional, you were taught a certain methodology for infrastructure, engineering, and operations that is wholly inappropriate in the cloud. 

Our approach is totally different. We do not view cloud as a hosting environment. We view it as a platform that you build on top of. And ideally, you build as high up the stack as you can, serverless being the pinnacle, but maximizing the use of platform services and minimizing the use of infrastructure as a service, is the most important concept. And that is really hard to scale because there’s not enough talent with that specific skillset and experience.

And that’s really what we’re trying to solve for. How do we train people either mid-career or earlier in their career, maybe as early as intern level, to build on top of cloud platform services and to avoid as much as they humanly can as VM hosting, VM snapshotting, VM cloning, the things that we used to do as the best available option at the time, but which we now know only perpetuate problems, especially from a security and disaster recovery standpoint.

So if you can’t provably recreate your infrastructure through automation on zero notice, you fail the fundamental DR BC test. If you can do that through cloud formation, through Terraform, through some deployment automation and configuration management infrastructure as code solution, you’ve solved not only your DR BC problem, but a whole slew of other problems along the way.

And so our idea is solve for the hardest part first. And if you do that, you probably get all the other check boxes checked along the way. And in this case, I think the hardest thing is can I provably recreate the running system state, at any moment, with no hesitation and no fear of meaningful data loss or fumbling or whatever else? That’s the litmus test, as far as I can tell.

Rahul Subramaniam: So the folks that you’ll be hiring into Cascadeo University, are they primarily computer science college grads, or do you also look for talent from other non-traditional educational backgrounds?

Jared Reimer: I think maybe I’m a little bit biased because I have an academic computer science background. I think for us, understanding how things work and why they work is very important. You can teach people how to pass tests. You can teach people how to do repetitive, manual, kind of labor operations, but that’s not the same as teaching them to think.

And I think the advantage that, at least for me, the thing that I got out of my computer science background, is a fundamental understanding of why things are the way they are, why they work the way that they do, how they work under the covers. And I place a pretty high value on that, personally. I think that’s really important.

Some of our people don’t have CS degrees, but it’s a pretty important thing, I think, from us from a recruiting standpoint. The other thing is the advanced AWS certifications, like Pro Architect, are so difficult to attain that you can’t really memorize your way through it. You can memorize a bunch of facts and figures, but the exams are structured in a way that make it almost impossible to just take the test and pass.

Amazon has consistently insisted on the highest possible bar, and it’s a serious thing. We just went through the MSB audit again, and it’s a massive amount of work. It’s very intensive. It’s very in depth and rigorous across a whole range of disciplines, and their technical training and certification is the same. So we really think that’s a shining strength of AWS.

Hilary Doyle: Cascadeo stands out as an innovator in hiring and training. I’m sure. Jared, you’ve looked across the expanse of other companies looking to solve this talent shortage problem. Are there any failures that you’ve seen, that come to mind?

Jared Reimer: I think the failure is mistaking cloud for hosting. And that’s very, very, very common. And sometimes it’s intentional because it scales. It’s a much easier thing to scale VM hosting on a cloud platform, than it is to build .NET native.

But in the end, the only real winner in that is the cloud provider. It’s certainly not good for the customer. It’s not good from an InfoSec standpoint. It’s not good from a DR BC standpoint.

It might be sort of good in the short run, for the professional services or consulting partner, but it’s not good for them in the long run either. It’s a short-sighted shortcut or Band-Aid to a fundamental scarcity problem. And I understand why much larger consultancies do that. They have to. They’re doing it out of necessity. They’re not doing it because it’s what they want to do, necessarily.

But if your goal is to have a 100,000, quote/unquote, “AWS experts,” as some have claimed, there’s no other way to do it. You literally have no alternative but to dumb the problem down as much as you can and simplify it and make it repeatable.

And if you’re trying to build McDonald’s, then great. But if you’re trying to build a Michelin-starred restaurant, that’s not how you do it. And of the two, I understand that McDonald’s is way more lucrative and scalable and repeatable. And McChicken tastes the same everywhere you go in the world.

Hilary Doyle: But you’d rather be at the French Laundry?

Jared Reimer: It’s uninspiring. I would much rather do something small and significant and work with really interesting customers.

Hilary Doyle: Yeah.

Jared Reimer: And I think our engineers appreciate that. Again, we’re we’re not looking to scale like mad. We’re looking to scale through excellence and through innovation and through competent delivery of cloud native solutions.

Rahul Subramaniam: Jared, one last question from me. Who, in your opinion, has the upper hand in acquiring top talent right now? Is it the small niche players, who are so differentiated? Or is it the big players who have the brand recognition and deep pockets to kind of go after all this talent?

Jared Reimer: I think the right talent, the people that love the game and love the technology, probably are more inclined to be attracted to smaller, more nimble, agile, leading edge, like working with the latest technology, if only because they can move faster. They can do things that the bigger players have to have committees and budget discussions and planning meetings and whatever else about.

Rahul Subramaniam: Got it.

Jared Reimer: We like to be in the middle. We’re somewhere between the two, and I think that’s actually not a bad place to be.

Hilary Doyle: It’s a great perspective to go out on. Thanks so much, Jared.

Rahul Subramaniam: Thank you so much, Jared. This was excellent.

Jared Reimer: Absolutely. Thank you for having me. I really appreciate it.

Hilary Doyle: Rahul, listening to Jared talk about his struggles to find and train cloud talent, it just sounds like we’re living through an Industrial Revolution.

Rahul Subramaniam: Yeah.

Hilary Doyle: It’s shifted our work paradigm and entrepreneurs and business owners are going to have to manage their way through a skills gap, while the rest of the world catches up. We’re going to catch up though, right?

Rahul Subramaniam: I’m pretty confident we will, Hilary. I mean, just think back to the First Industrial Revolution. I mean, mechanization was spreading so fast and wide, that talent that was trained for the factory was nearly impossible to find. I mean, it took a while, but what followed, changed everything. 

Our entire education system is a product of trying to get a workforce ready to be productive. We now look at our economies in measures of industrial and individual productivity. I’m quite certain that this cloud-centric revolution, that we are all a part of, is not just going to change how we do things in the future, but it’s already changing our present.

I mean, imagine what the pandemic would have been like if we didn’t have all the technologies, and here I mean cloud, that made discovery of medicines possible so quickly. Or on a more mundane note, made remote work and online retail possible, that allowed us to survive through this pandemic. So again, I’m an optimist. I think we will not just figure it out, but we need to figure it out because it is a global imperative.

Hilary Doyle: Before we move on, if you are listening to this podcast thinking, “I could really use a deep dive into the technical weeds of cloud computing,” well, Rahul has been moonlighting.

That’s right. He has a live stream with a different co-host. Is that co-host as charming and witty as this co-host? Well, yes. Yes, he is. More so? Probably. But does he have my facility with puns? Absolutely not. Here I have a clear and insurmountable advantage.

Rahul Subramaniam: Hilary, no one has the facility that you have with puns.

Hilary Doyle: Isn’t that just the truth?

Rahul Subramaniam: And I love you for being my co-host on this show, but I do have another show. Every week, I’m joined by fellow CloudFixer, Stephen Barr, for a livestream on all things AWS. It’s called AWS Made Easy and you can find out more about it at CloudFix.com/livestream.

We stream live on LinkedIn, YouTube, Twitch, Facebook, as well as Twitter. You can ask your questions live on the show, and we, along with our amazing guests from AWS, will do our best to answer them for you.

Hilary Doyle: Well, that sounds both great and also incredibly helpful. Also, who doesn’t love Steven? So proceed with the side hustle. Rahul, you have my unwilling blessing. You have my casual blessing.

Joining us now, to talk about big picture, the massive picture, when it comes to hiring cloud talent at huge scale, is David Brassor. He’s a managing director and partner at Accenture.

David, we are delighted to have you. Let’s start with some background. Accenture launched Cloud First in 2020, with a three-year, $3 billion investment, to enable, accelerate and implement cloud adoption.

The cloud practice was already pretty significant at Accenture. I think it brought in $11 billion in 2019. You are leading the Cloud First design practice. So practically speaking, can you tell us what that means?

David Brassor: Sure. So if you think about Cloud First design, what Cloud First design is focused on is think about the tip of the spear for organizations looking at cloud adoption. So everything from their initial cloud strategy, cloud reference architectures, roadmap, there’s a number of things that we’re seeing across our clients, and one of them being they’re not seeing the benefits of cloud that they expected.

So in many cases, organizations took, I’ll call it early and fast leaps to cloud. They lifted and shifted existing workloads to cloud, and they’ve now since figured out that they’re not getting the value from the lift and shift versus the broader cloud transformation opportunities around modernization of legacy workloads, et cetera. But more importantly, tying that back to business value around cloud adoption.

So many of the organizations took a pure technology-oriented approach to cloud adoption early on, and now they’re finding they’re not seeing the true business value of that cloud adoption.

Hilary Doyle: That’s a great segue because we’re looking at the ways in which we hire and structure teams for the cloud. We’re thinking about ways in which analog can inform cloud. And when it comes to hiring and building out teams, how are you instructing your clients to look at their teams for 2.0?

David Brassor: Yeah, a big part of that is re-skilling their existing talent structures.

So one of the things that we’ve seen is many of our clients are struggling to go and find that talent, even with a marketplace that is seeing layoffs across some of the major cloud providers, working with a client right now who’s been working for months and months to hire for a senior cloud resource.

I look at some of my other clients who are, in some cases, looking at having to hire hundreds of people across various cloud capabilities, and they’re struggling to find the talent in the market.

Rahul Subramaniam: So David, in the market, we actually see so much noise about not finding the right talent, but could you articulate for us what it means to be skilled for the cloud, in particular? And how does that compare with the skills that they possess today?

David Brassor: Yeah, so if you take a look at a lot of organizations, they’ve got people who have had gone and done their cloud certifications. They have some basic understanding of cloud and cloud capability.

And then in many cases, they’re taunting, “I’ve got multiple certifications across multiple cloud providers.” But when you get down to have you been there? Have you done it? Have you got the t-shirt that says you’ve done it? The answer, in many cases, is no.

I may have theoretical knowledge, or I may have some initial skillsets and capabilities that I’ve been able to leverage on the cloud, but when you actually look for resources that have lived through large scale cloud migrations, that have lived through cloud ops, that number of skilled resources goes down, exponentially. And that’s what our clients are looking for now.

Rahul Subramaniam: So David, in the first wave of these migrations, we were all DIY. Companies and organization took whatever skills they had and whatever resources they had, and they just went and did the stuff. And in the process, they learned.

What’s changed now? Because we still have less than 10% of the global workloads in the cloud. And of course, yes, there’s a huge demand to move there very quickly. But what’s changed between the first wave and this second wave?

David Brassor: So I mean, the first wave was the easy stuff, as I’ve always put it. So this was organizations who looked at simplistic workloads that I could lift and shift into the various cloud providers.

But now, what you’re seeing, is they organizations are tackling the tough things. They’re tackling automation. They’re tackling containerization of workloads. And now, the skillset requirements have grown exponentially. But in many cases, unfortunately, the talent hasn’t, in that people don’t have that experience within many of our client organizations.

So that’s why you’re also seeing them reach out to a lot of the big SIs to say, “Hey, how can you come help me? Because my organization just doesn’t have the talent base.”

Rahul Subramaniam: Yes. And I see this as a big conundrum. On one hand, you have this massive dearth of particular skills, which is cloud technologies. The number of people you have who are truly experienced in migrating complex workloads or even running them in the cloud. Is a finite number. And of course, a limited resource.

Given that there’s a logical argument to be made that you would take the existing resources you have and train them as fast as you possibly could, to get them skilled in this new technology, for which there just seems to be infinite demand, why then are we seeing massive layoffs? Especially in engineering when they could be retrained or reskilled for the cloud? Because well, everyone is either in the cloud or moving to the cloud.

David Brassor: So I mean, it’s an interesting comment, in that one of the dynamics that we’re seeing is that they’re actually not spending a lot of time on the market. They are, in fact, being snapped up by other organizations. And in many cases it’s organizations that are going through their cloud journeys and looking for that talent.

And we’re seeing some of the SIs snapping up some of that talent also.

And for a lot of these organizations, it’s incredibly important that they start to think about how do they actually retool their existing people.

Hilary Doyle: I’m assuming that you’re working with large and small companies at this point, David. Can you give us an example of someone who has done this well? Surprised even you?

David Brassor: If I take a look at some, one organization that we’ve been working with, they were very early adopters to cloud, but they retooled their existing organization. They leveraged SIs, to accelerate their journey to cloud, and made sure that there was a proper knowledge transfer program in place. So their team eventually was brought along for the journey.

So as they built skillset and as they built competency, they will then transition some of those next steps of their cloud journey to their own internal teams.

But they were also, along the way, recognizing where’s the value-add of leveraging a third party service provider, and where should I be leveraging my own teams internally, moving forward?

So making that decision around what skill sets and capabilities do I need to bring in externally versus what things do I want to build on internally, for the long-term success of the organization?

Hilary Doyle: So that’s a great answer, but I also recognize you’re a consultant, and that is a very consultancy answer that you’ve just given us.

So I want to understand, I mean, you’ve laid out all of the points, and yes, of course, makes perfect sense, that’s how you should evolve into the cloud. But what is the timeframe we’re talking about? I mean, that speaks to such foresight, and a lot of companies don’t have that foresight because they don’t have the T-shirt yet.

David Brassor: Right.

Hilary Doyle: So help us understand, really practically, how this is happening well and how it is failing spectacularly? Because I know that you’re seeing both sides.

David Brassor: Yeah. So I mean, how it’s happening well, is that organizations are taking a bottoms up approach to building cloud skillsets, capability, and competency.

So they’re setting the foundations appropriately. They’re training their people, going out and building the landing zones, integrating with security, doing all the right things from a foundational perspective, building automation, DevOps, DevSecOps, into their strategy early on and Agile, and then building on top of those core competencies and capabilities.

Where we’ve seen organizations fail is when they attack cloud, as the old saying goes, willy-nilly, where they’ve got competencies going and programs going on here and programs going on there, and no alignment across those initiatives. And those are the ones that, in some cases, unfortunately, we typically will also see that they run into security issues and risks and scenarios, where there’s been a breach because somebody made a cloud configuration error that caused the environment to be exposed.

Rahul Subramaniam: So David, I have a follow-up question on this. From everything that I’m hearing, it seems like there is a need for foresight planning and, of course, the background to make all the right decisions, as you go down this path. What are those basic skills that they need, even to just get started?

David Brassor: Yeah, so even to just get started, if you take a look at where organizations have been successful in this, it is bringing people into the organization who have some experience, who’ve lived through prior cloud adoptions, understand the hurdles and how to address them.

So I take a look at one client that I’ve been working with who is on their cloud 2.0, I’ll call it transformation, and they have surrounded their team and brought external talent in that have experience and have been through this journey and know what things to do around setting up your operating models, your governance structures, workload assessments, et cetera, but then they’re also bringing their internal people along and building capability there.

Hilary Doyle: For any young developers who are listening to this and feeling sad, how would you suggest they focus as they try to enter the workforce?

David Brassor: Yeah, I mean, I think cloud native development continues to be a significant requirement for pretty much all of our clients who are now looking at their existing legacy workloads, that maybe legacy code bases like .NET and others, and they want to move those to cloud, so they are continuing to have the need for skillsets out there that can develop in cloud native capabilities across the various cloud providers.

So my perspective is if you are coming at this as an application developer, cloud native should be a significant focus area for you, moving forward, and you will become a valuable asset in the market if you can showcase cloud native capabilities.

Rahul Subramaniam: Got it. I have asked this one question, which is, if you could hire just one person on your team that is responsible for kicking off cloud transformations, who amongst everyone out there would you pick as your first draft?

David Brassor: What I’m constantly looking for, is people that have lived through complex cloud transformations, have been through not just lifting and shifting existing workloads, but going through application and infrastructure modernization, have gone through containerization, have gone through cybersecurity integration and built fault-redundant solutions in the cloud. So somebody who’s really got that experience of saying, “Hey, I know how to architect for cloud. I’ve done it. I know where the pain points are. I know how to address them.”

And I’ll be honest, that is an incredibly tough skillset to find.

Rahul Subramaniam: Any specific name comes to mind, amongst all these celebrity rockstar cloud transformers?

David Brassor: I’m not going to name names. I’m just suggesting it is more of a capability piece.

Rahul Subramaniam: For the record, I pick Werner Vogels. If I had him on my team, I’m all set.

David Brassor: That’s a really good choice.

Rahul Subramaniam: So David, it looks like if I just look at the entire resource pool out there, we’ve traditionally had very few women in tech, but it feels like the cloud is one of those paradigms which levels the playing field.

Everyone is rushed at the beginning, it seems. Everyone is learning. Everyone is trying to figure out how to make it work. How are you seeing that transformation play out, when you look at the resource pool, especially when it comes to gender equality and the cloud?

David Brassor: Yeah, I mean, I think cloud has been the great leveler of the playing field in that aspect, in that, as you said, there is such demand for the skillsets out there. So we are actively seeing, we go into many of our clients and in some cases it’s a 50/50 mix of female and male talent within a number of their cloud environments.

So you’re seeing a number of female tech executives, female tech senior leaders, playing senior leadership roles across many organizations. So a great example, one client I’m working for, their SVP that has responsibility for all of their IT and operations is a female resource. And we’re seeing that across so many of our clients that, ultimately, the cloud has really been a leveling of the playing field for female talent in that it has allowed so many people to jump into senior leadership roles.

And I think across so many of my clients as well as some of the software partners that I work with, I get on calls and in many cases, and a great example, I was on one yesterday, where there was more female cloud talent on that call than there was male.

Rahul Subramaniam: What is the one piece of advice that you have for organizations that are charting out this cloud journey, from a talent perspective, that they should absolutely not do?

David Brassor: I think what they should not do is, I guess, is sit on their laurels and assume that I’ve got some people that are helping me with my cloud migration and cloud transformation. They need to continue to look at the marketplace and look for new skillsets and capabilities as their capabilities evolve.

In many cases, as I said earlier, so many of the cloud people within many of our client organizations are carrying and wearing multiple hats. They may be doing cloud ops. They may be doing architecture. They may be doing security. And as you grow, that’s not a sustainable model anymore. I need to build specific competencies and capabilities. And so a great example with one client, we had a client who had a major breach, a number of years ago, and they attributed it back to the fact that they had a very small cloud team that was so diverse and was wearing so many hats, that they started to not have enough depth and experience to be able to ensure they were properly securing and controlling their environment.

And ultimately, that was what they determined, that led to a breach within their environment, was not having that, enough people with the right level of depth and expertise, to make sure that they weren’t going to end up on the front page of the press because of a breach.

Hilary Doyle: David Brassor, thank you so much for your time. Thank you.

David Brassor: Thank you.

Rahul Subramaniam: It was a pleasure having you, David. Thank you so much.

David Brassor: Thank you for having me.

Hilary Doyle: David Brassor, sharing insights from his extensive experience in the cloud. I wanted the two of you to go head-to-head, but everyone seems to agree on the premise here.

Plus, David is Canadian, so it was bound to be a pleasant conversation. But what are the Subramaniam hot takes to what was just said?

Rahul Subramaniam: So the conversation really made me even more convinced that expecting to find experienced resources who have mastered cloud patterns is really a fool’s errand. I mean, number one, there is this super tiny pool of resources who would qualify against that bar, and most likely, they already work for one of the cloud service providers.

Hilary Doyle: Right.

Rahul Subramaniam: Second, is that the cloud environment is changing so rapidly that keeping up, in itself, is a mammoth challenge. I mean, this is a continuous game of catch up. And in this particular case, you really need to rethink what it means to hire talent.

Hilary Doyle: But what do you mean by rethink talent? I mean, do we have to think differently about when and where we hire talent? Is there any single fix to this problem of finding and leveling up your talent pool?

Rahul Subramaniam: So there’s this two-pronged approach that I would recommend.

The first is that a lot of organizations are looking for the top 1% talent pool, but limiting themselves to a particular geography. If you hired from the top 1% globally, instead, you just have a significantly larger pool to recruit from. I mean, that’s just how math works.

Hilary Doyle: The wrath of the math. I like it.

Rahul Subramaniam: And the second is that you might want to augment your talent acquisition strategy with talent development. I mean, this is what we heard Jared talk about today. Training your staff for the skills and tasks you need them to perform seems like the need of the R.

And I’m hopeful, though, that this would just be a short to medium term Band-Aid. The reality is that there’s this huge gap in supply and demand of talent, but the way capitalism works is that the wider the gap, the faster it gets filled.

Hilary Doyle: Sure. I mean, it doesn’t sound like that’s been the case, to date, but maybe that starts changing now. There were different opinions between Jared and David when it came to re-skilling your current talent. David was saying, sure. And Jared was really saying old dogs, you know the adage. So what’s your opinion?

Rahul Subramaniam: I think I would side with Jared as default because the cloud is a completely different paradigm and needs this hard reset in thinking.

Hilary Doyle: Mm-hmm.

Rahul Subramaniam: However, if you have high caliber resources that can learn fast and change their mindset about how things are done, I can see re-skilling being a viable option. I think the only advice I can give someone is something that we follow ourselves. First, hire smart people. People who can learn fast.

Second, kickstart your cloud projects quickly and create an environment where folks can experiment, fail, and learn and deliver solutions quickly. I think this really is the fastest way to build a cloud-centric organization. 

Hilary Doyle: So we’ve covered fast, but what’s the simplest solution to cloud staffing challenges?

Rahul Subramaniam: Use automation and tools to do everything that humans find mundane or repeatable.

Hilary Doyle: Bam. How long until this actually levels out and we catch up? Does that ever happen? Or does this challenge just become more entrenched over time?

Rahul Subramaniam: Hilary, as you know, I’m an optimist.

Hilary Doyle: So am I.

Rahul Subramaniam: And the reason why I’m an optimist is because there are two invariants that I know of. The first one is that AWS and the move to the cloud is an imperative for business. It’s already happening. The second one is that with all the amazing work that AWS is putting into certifications and education, the pool of potential resources that is starting to get hands-on experience is growing at an incredible pace. So the gap in cloud knowledge is really shrinking fast. The bottom line really is that the future definitely looks brighter and better.

Hilary Doyle: Okay. It’s a huge conversation. We want to keep it going. So you can email us at podcast@CloudFix.com. Please send us your thoughts about the show and what you’d like to hear about in future episodes.

Rahul Subramaniam: Also, please tell everyone how you feel. Leave us a review, and don’t forget to follow the show to get new episodes as soon as they drop.

Hilary Doyle: AWS Insiders is brought to you by CloudFix. They’re an AWS cost optimization tool and you can learn more about them at CloudFix.com.

Rahul Subramaniam: Thanks for listening.