Ep. 19 How Are AI and Costs Reshaping Cloud Transformation? with Shadman Syed

Matt Pacheco
Hello everyone. Welcome to the Cloud Currents podcast, where we explore the strategies and technologies powering innovation in the cloud. I'm Matt Pacheco, your host, and I oversee the content marketing strategy at TierPoint, a managed cloud and data center services provider. Our guest today is Shadman Syed, the director of cloud services at Paradigm. With over 20 years of IT leadership experience, Shadman has been at the forefront of cloud adoption across various industries, from manufacturing and software development to cutting edge tech like IoT and augmented reality. At Paradigm, he oversees the architecture and operations of their enterprise cloud platform. In our discussions today, we'll dive into strategies for successful cloud migrations, talk a little bit about scaling globally distributed environments, optimizing cloud costs, ensuring business continuity, and adopting new technologies like AI. So with that, thank you for joining us today.

Shadman Syed
Shadman, glad to be here. And Matt, thank you for having me.

01:19 - Shadman's Career Journey

Matt Pacheco
So we'll jump right into it. I'm really interested in understanding your career journey and what sparked your interest in cloud computing. So can you give us a little bit about your background?

Shadman Syed
Sure, sure, we can start. So right now, as you said, I'm director of cloud services at a company that specialize in developing software and building infrastructure industry. As paradigm here, I am responsible for the architecture, engineering and support of our enterprise cloud platform. So my company develops cloud based products and services for its customers, and my team and I, we host the product as software as a service in the cloud. So, briefly about myself, I have several years of experience working within the IT industry. However, my focus over the last decade or so has primarily been on designing, engineering and supporting products and services in the cloud. Throughout my career, I have held various roles in managed services, shared services, and professional services with organizations of all sizes.

In these roles, I've had the opportunity to implement and manage cloud solutions that have been tailored to meet each organization's unique requirements. So, going back to the beginning of my career, I have bachelor's and master's in electrical engineering. After completing my master's program, I started working as a systems engineer at a multinational IT consulting company, an ORc company in Michigan in the suburbs of Detroit. In that role, I was fortunate enough to gain valuable experience in hosting products and traditional infrastructure such as servers, storage units, network equipments, all hosted within the data center. After that, I worked as a systems engineer at an automotive manufacturing company and then a telecom services provider. There, I started to move up the food chain and became a senior systems engineer, then a lead, and finally as a manager.

03:22 - Transition to Cloud Computing

Shadman Syed
After that, I made a major career change and moved from suburban Boston from suburbs of Detroit to Boston and join a french multinational company that specialized in IT consulting and outsourcing. This is where I first got my exposure to cloud, on which I'm going to talk briefly in about a minute. So after that I got my first director level position as a director of cloud services at a software development company in Boston. And followed by that I held different position as director of IT infrastructure, cloud DevOps, cybersecurity, and few other similar roles in small to mid sized companies. Which finally brings me to my current role here at Paradigm. So my first exposure to cloud happened about eleven years back when I was working at the french multinational and then I was involved in a project for hosting an SAP ERP system in AWS.

Back then there were not that many companies who hosted SAP and cloud, as most companies only knew, and they were comfortable with hosting SAP on premises. So from that point onward, my career in cloud just took off. And since then I have been engaged in a number of cloud migration initiatives, cloud improvement assessments, application refactoring in cloud using cloud native services, and then also leading teams that specialize in cloud architecture support DevOps and lot more. That sort of summarizes my history so far.

05:01 - Balancing Innovation and Operations

Matt Pacheco
Wow, you've quite the journey. I mean you've touched a lot of different things there from all of the, I guess, experiences you had. How have your perspectives evolved in balancing like the innovation and operational realities when it comes to adopting emerging technology? So you've been introduced to new technologies throughout your whole career. How has that affected you?

Shadman Syed
Affected me and actually my team? In many ways it has changed me a lot. I mean, as I said, I started my career in traditional infrastructure management on premise, moving to cloud and adopting cloud native services, using infrastructure as code. Even on when I was working with the traditional infrastructure on premise, we did automation, it was through. Usually my background is mostly on the Linux side using shell scripting. There was still automation there, but shifting to traditional software development type of infrastructure as code, where you have coming up a new feature, new config, you create a branch, commit your changes, merge it, do a pull request. That concept was foreign to me and that is foreign to also traditional infrastructure engineers. And that's the biggest struggle that I have seen.

That moving from most of, I would say click ops type of work to DevOps. Okay, that has been, I would say the biggest struggle once I've got it. Unfortunately, there are not that many people who explain it to you that this is how it needs to and this is sort of the benefits and advantages. Most cases people, they just learn it on their own. But I wish that somebody or course of my career has told anything. Moving to cloud from traditional infrastructure, automation, using infrastructure as code, using cloud native services, managed services, whether it's for database, for caching, redis, or so on and so forth. This is where you get the biggest bank with your buck. That's, to answer your question, has been, I would say the biggest thing that I've learned so far.

07:24 - Cloud Migration Challenges and Strategies

Matt Pacheco
Excellent. And it sounds like earlier in your career you were migrating from on premises infrastructure to the cloud. Not all businesses are quite there yet. There's still a lot of businesses still doing that today, specifically going all in, trying to move it to the public cloud. So can you tell me a little bit more about what have been some of the key driving forces behind large businesses and enterprises deciding to go all in on public cloud adoption?

Shadman Syed
Multiple reasons. In many cases I've seen is this modernization? There are certain things that you just cannot do on premise, that you can either just do it in cloud or easily do it in cloud. Okay? Cloud just gives us the cloud native services that I've talked about. Not just that the serverless compute, so on and so forth. They're just not there on premise. Even if you have certain applications that are even using microservices based architecture, the backbone for that is containers, the container orchestration or its management on premise, there are ways to do it. This is mostly manual or through kubernetes, which is quite complex. Cloud also gives you even container orchestration and other services, native tools that you can just easily adopt and move forward. Biggest challenges that I've seen with the customers is that first of all, migration to cloud.

If you have not done it before, don't try to do it on your own. Partner with someone. All big cloud service provider, AWS, Azure, GCP, they have partners that will help you with that migration. Start small. Okay. And they would also sort of give you the benefits for just moving your infrastructure to cloud, your application. Certain credits there also, okay, so partner with someone, start small and do an assessment that these are certain applications. Frankly, not all applications are also designed. If you have something legacy that was developed 510 years back and you don't even have people who have developed it even around probably it's not a good idea to just migrate that as is, you're going to think about either replacing it or refactoring it when you work. So that work by itself requires an assessment.

Doing that on your own, there's a lot of challenges. Partner with someone who has been on this journey before and can help and guide you with that. Once you have done this a couple of times and you have moved certain application in a group, then going forward you will develop that confidence and that expertise and then you can just think about moving, you know, of your workloads in cloud. That's the biggest challenge that I have seen so far with the companies. And my biggest advice there is, if you think about moving cloud, think about the right reason why you're moving it. It shouldn't be just, hey, cloud means modernization. In many cases, that might not be the case. Partner with someone who has done before, who's been on this journey before.

11:20 - Cost Considerations in Cloud Adoption

Matt Pacheco
Excellent advice and I parrot that advice. Work with the manage provider. Hint, hint. So were talking a little bit about migrating to the cloud and re platforming and refactoring, all that. What are your thoughts on the trend of some businesses doing the opposite and repatriating from the cloud? I'm curious to hear what your thoughts are on that.

Shadman Syed
Repatriating meaning just going back from cloud to on premise?

Matt Pacheco
Yes.

Shadman Syed
I mean, there are certain reasons. What I have seen in some cases, one of the reason is security and compliance. Cloud is by nature is shared infrastructure. You can still have certain options. We can get dedicated hosts, you know, things of that nature. But overall it is shared infrastructure. Companies who have strict, some like financial and healthcare companies who are really concerned about the privacy of the data, maybe it's better for them to get that peace of mind to host in cloud. Host on premise instead of cloud. That's one thing. Second thing is also is cost. We don't factor in costs. In many cases we think cloud is on demand access. But if you have resources just sitting around orphan or you have oversized, underutilized resources, your cost is going to go up.

And if you have not done, if you don't have a good financial office, I call it the fin offs team. Eventually, at the end of the year, if your budget is not in alignment what you forecasted, someone is going to come and knock at your door and say that, hey, your cost has gone up. Okay? And that will be a tough conversation. Okay, so cost is another driving factor because on premise, once you have paid for your infrastructure, okay, it's just all free, okay, it's just after that you just run your whatever the cost is for your facilities, for power and cooling, so on and so forth. Cost is another reason. Other thing also is it's not in my field, but I hear it a lot with this adoption of AI and ML services.

Some of it requires higher power like GPU's, not standard computers, not standard cpu's. And running those workloads in cloud again is the cost factor. It can get really expensive running those workloads on premise and then being little like control freak about that. Hey, my AI models are just totally in my control. They're not in a shared environment. That's another driving factor. So security and compliance cost and some AI workload, those are the three places where I've seen that some companies have gone back from cloud to on premise.

Matt Pacheco
Excellent. And since we're on the, we talked a little bit about cost and jumping ahead of my plan here. But back to the cloud migration piece and migrating to the cloud. How do you approach cloud cost optimization and ROI calculations? When advising enterprises on their cloud journey.

Shadman Syed
Then number of assessments similar what would be my cost? So first of all, you have to know what your current cost is on premise. You have to calculate everything. If your hardware is not fully amortized, what's the remaining costs? The cost for facilities, power, cooling, people who are managing the infrastructure, all of those things, then there are certain tools out there industry. I don't want to just name the names, but if you just search, if you install them on premise, there's an agent base with an agent list that will collect information on premise, and we'll see that for your current infrastructure. Are you oversized or undersized? If you'd work, basically move the workload to do cloud, the rightly sized solution, how that would look like bringing all those pieces together, doing an assessment for your current cost on premise.

Working with these requires multiple workshops with the right stakeholders who know their on premise infrastructure quite well, setting up certain tools who gather that data, that insight on their current setup. And then if you run similar workload on cloud, you just come up with the right cost model that this is, if you move these applications, okay, to cloud, this is how much it's going to cost you. And then you do an Apple to Apple comparison. Other thing with cloud that you, the way you reduce your cost in cloud is it's not just the on demand cost for the infrastructure. If you, if you have, if you think about that, you can use reserve instances, typically reserve instance, you can go with one to three years and it gives you significant saving, anywhere from 30% to 60% to 70%. Okay?

Then on top of that, certain cloud service providers, they also give us savings plans on top of it. Okay, then if you commit, if you know that your cloud cost is going to be, for example, 5 million per year, if you commit that you're going to spend 5 million, they will give you additional credits. So there are ways that you can further reduce your cloud costs. Bye. Using reserve instances, saving plans, working with your cloud service provider, committing to a certain number of cloud spend and getting additional credit. That all, I mean, what I just mentioned is a lot, but it just factors into the whole discussion about rois, cost comparison and all.

And if you, again, as I said in the beginning, if you partner with the right service provider, they will give you that information and then you can make like a more informed decision based on all these factors.

Matt Pacheco
So are there any other certain cost models that drive more savings than others? Are there? So I'm looking to adopt private public cloud. Are there certain things you mentioned reserved instances, are there any others that people should consider?

Shadman Syed
Yeah. So there's depending on your workload, if it's not production and it is, you're doing a PoC or half hour, some high computational workload. Spot instances is another way. Think about that's literally fraction of a cost of your on demand. Okay, just keep that in mind with Spotify, the cloud service provider can just take that access away. If they have an actual paid demand compared to spot demand, people watching this podcast would know that they will shut you down and go with the paid subscription. So spot instances in another way, but it's not designed for production workloads. And I know that many companies, their biggest cost driver is not just production, is they are doing number of tests, pocs, things of that nature. Spot is also a good choice for them.

Matt Pacheco
Excellent. Interesting question. What are some of the things you've come across or some of the things that customers didn't expect when going into this? Yeah, sorry.

Shadman Syed
Yeah. Okay, so you're saying that if moving workloads to clouds, what has been the biggest surprise for the customers?

Matt Pacheco
Yeah. For. As it relates to cost too.

Shadman Syed
As it relates to cost? Yeah, as it relates to cost, yeah. It is in many cases what I've seen that they're, again, it goes back to the initial assessment. Many environments are oversized. And if you move an oversized environment to cloud, if you utilizing it just ten to 20% and you have like this much resources allocated, that's a waste of resource. So the biggest thing that you should win, that's the beauty of cloud. If you have set it up right, you can easily scale up and scale down your resources based on your workload. Good example is I have worked, I mean, again, probably should not name with some big retailers as our customers. They have Black Friday their workload on that Black Friday or that weekend just skyrockets. It's not even call it me. It's like exponential, like a hockey stick.

It just goes up on premise. Most customers, they will have an environment sized for that workload most of the year and they use it only for three days or maybe a week. If you run the same workload in cloud and you have proper auto scaling setup, that's where you would save a lot. Other thing also is sometimes we've seen that you have these orphaned resources, for example, backups, your database backups and is in terabytes or something. And then if it's just sitting out there, that is no longer needed that again, I've seen that people get all of a sudden like a sticker shock at the end of a month. And to see the bill that, hey, this is, I don't know what this resource is. This backup is like ten terabyte, like some orphaned resource sitting out there.

Those are some of the surprise elements that I've seen.

23:57 - Infrastructure as Code (IaC) and Its Benefits

Matt Pacheco
So aside from cost, what are the other surprises that customers or businesses find when they're adopting the cloud to broaden it just a little?

Shadman Syed
Yeah, yeah. It is a change in scale. Okay. There are a lot of similarities between on premise managing your infrastructure on premise then in cloud also. But there are a lot of differences too. Okay. The true benefit, as I said, you'll get out of cloud is by automating as much as possible from your infrastructure build from smaller changes to literally new employment deployments, application, you know, the continuous integration and continuous deployment CI CD pipelines and all that. Skill gap is one thing that I have seen that it has been a big surprise. And I'm just trying to think what else. To me, the biggest thing that I have seen is the gap in skills knowing about the cloud native services, taking advantage of that.

Some organization, in fact, most organization initially when they migrate to cloud, we call it like a rehost or it's a fancy name for lift and shift. That's not the right way to run it. If you run the same workload as is on cloud, you're going to fall behind a number of places. Cost is going to be one thing and you won't be able to modernize your application that much. To modernize, you're going to use cloud native services. Think about if you have more of a monolithic architecture, go with a different microservices base. Okay. Another thing that I have also seen is in some cases you may see a performance degradation because now there's some network latency involved.

If you have your user base close to your site, your on premise location somewhere in east coast, and for whatever reason you have decided that you're going to host in central region, that latency is going to play into some performance degradation. So just be mindful of that. That where you're exactly hosting where your user base is and there is a way to get around it also using some CDN type of technologies and stuff like that. But latency, performance degradations, challenges related to that I have also seen and experienced to when companies, they move, they work.

Matt Pacheco
Very interesting. So I want to jump back a little earlier you had mentioned infrastructure as code, so I'd love to talk about that a little with you if you can explain what it is and also the benefits of it versus other models.

Shadman Syed
Yeah, so infrastructure is code. It's interesting. As the name says, we are provisioning infrastructure purely based on code. And being an infrastructure guy first. That sounds a little bit for I can improve an infrastructure by clicking certain buttons in a portal. So why do I need to code it? The biggest reason is if you code it is going to make your infrastructure faster to deploy. It will be consistent and it will be less prone to human error. The way I will provision an infrastructure compared to someone else could be different, maybe both right ways to provision, but there will be certain differences. The third person comes in and sees the same infrastructure, they'll see the differences and they might get confused. So the biggest advantage of that is make it consistent, repeatable and less prone to error. Okay.

Once you have your infrastructure in code, the nice thing is that it is if you're using a cloud agnostic tool, like for example, I'll use the name air terraform, it just declares that this is what this data I want. Once you have that code properly set, this just goes in and make sure that your state is consistent. If you ever get into a disaster situation and you lose your primary region, if you have everything set up as iacdez, you can use the same ise in some other region and quickly deploy it rather than now think about clicking this, that and the other. So literally a day long work now becomes few minutes. Like literally you just run a code and it just goes and deploys setups. Your doctor side, your infrastructure is ready. Rock and rolling there. Okay, that is sort of the advantage.

But the biggest challenge that personally I have experienced, my team has experienced is the skills gap in understanding why is it important? Start small, don't boil the ocean. Start with small piece of your infrastructure. Okay. Standardize. Use one tool. Prefer to use a cloud agnostic tool, because if you use a cloud sort of native tool, you know, aws, azure, they have their own IEC sort of tools that you can use and you're limited just to that cloud provider if you're using. Again, I'm not trying to promote any tool like you're using something like aerofone. You can just provision infrastructure anywhere, even on premise. Okay. Those are sort of the things that I've seen that always just celebrate success because it was just hit certain milestone. Make sure that you celebrate with your team.

This is the biggest place that I've seen that, you know, when it comes to ise, a traditional infrastructure engineer, an off version struggle. Okay. Help them get the right skills, start small, give them the right support, and celebrate success in the end, give them certain confidence and keep them on our lab.

Matt Pacheco
You sort of touched upon what I was going to ask you next, and it was what have been some of the biggest cultural and process hurdles as it relates to implementing infrastructure as code in organizations and teams you've worked with. You sort of talked about that a little bit. But I'm curious, are there any other big hurdles you had to get over?

Shadman Syed
The biggest hurdle, I frankly mentioned it, is about your background and this. The whole idea that I think I mentioned earlier in the podcast is even if it's not clickups, I can write this thing in a shell script or Powershell. What's the point in having this in a code repo and using sort of like same process that our application development folks use for if you need a new feature or even new or small config, basically you create a branch, commit your changes, do a pull request, have people like review your changes, approve it, and then merge it. This seemed like an overkill. No, this is the right process.

And once you get the hang of it, once you start using it, you will see that this dev process on the off site when you use it just brings wonders to your infrastructure provisioning and ongoing management. It reduces the drift, makes it consistent, less prone to human errors, and it's more of a training thing. And give your folks first a smaller task to learn this and help them out.

Matt Pacheco
What are some good resources you found to kind of stay on top of infrastructure as a code and any of that? Because I could imagine a lot of people listening here still trying to understand where to start. So it'd be really interesting to hear more about that.

Shadman Syed
There's a lot of resources out there. I mean, I don't want to again say promote anyone. But they're even simply on publicly available YouTube and all there are a lot of information out there. Okay. I frankly have seen is again, with this, for companies, it's better that again you partner with someone. Okay. In our case, we did partner with them. I'll be honest. We did partner with someone. Okay? And they brought in that, those best practices along with them and they told us why is it important? And this is how you use it. Why terraform compared to any other tool? Okay? And they help us to get on that journey. And we started small and we started growing slowly and now we are totally an IAC shop.

If someone, and we just jokingly say click ops and we just slap their hands, do not do it. If you do it again, I'm going to take your access away. Okay. Just kidding. But there are a lot of resources in public out there. Again, I want to just promote pluralsight has good training on courses on this. If you go on YouTube and all, there's, I've seen number of good training videos out for IEC. Yeah. Other thing also is that try it, I mean this IAC, you can try it locally on your laptop too. And then you have some small playground in cloud. You have your own test subscription, try it there. I mean that's how you're going to learn it. The best way that you do it yourself, do it locally, do it on your small subscription and cloud and partner with them.

Matt Pacheco
Really interesting. Another question I would have about that is selling the idea of infrastructure as code higher up. Like is there a special way to convince your leadership team to adopt this way of work?

Shadman Syed
It depends who is on your leadership team. If they are, if your leadership team is more it focused. They're not like, I mean, I'm just saying sometime it, you can have like a CFO, nothing against the CFO, but their role is totally different. Okay. They're, their priority is something different. I have not seen, frankly adopting, modernizing anything for improvement, any pushback. Most places that I have worked, depending on the, what their background is, if they don't have the right background in cloud or application development and all. So it may take, I mean this part of your, as part of the leaders also my role or people in my role. I mean, you have to not just manage down or across, you have to manage up. Also what I just told you about, you know, what are the benefits for it?

You gotta sell it to them. Okay. And in most cases I find that if they're reasonable. And if they don't find it like, hey man, this is gonna increase our cost too much, or this is a drastic change. And again, making these changes in small iterations, they will buy into it.

Matt Pacheco
Excellent. Are there any security concerns with infrastructure as a code? Any security considerations?

Shadman Syed
Absolutely. You got to make sure that you have site right security scanning, both for your static and dynamic code, also going on as part of your code repos. And then once even you have made your changes, committed your changes, and then let's say you deploy it in your non production environment. Production, there are a number of tools out there who real time scan your systems and look for any vulnerabilities. Okay. And also you can just do some monthly additional scans and will like a pen test and stuff like that. But what I have seen that, again, I just did number of tools that comes to my mind. I don't want to promote any tool that they do static and dynamics code scanning.

And then you could check your code repo, literally, and see what codes you have, and they will look for any vulnerabilities and issues and they will let you know. And then as part of your QA or your test, that you can make sure that whatever security team or that the tool has identified, you address it before you deploy it in production.

Matt Pacheco
Excellent. And I guess now to make it a little more complicated with all the things we talked about, because we talked about cloud cost optimization, we just talked about infrastructure as a cloud migration. How does AI play into this? That's a big topic right now. I would love to hear your thoughts on how AI integrates. We can start with how does AI integrate with infrastructure as a code, if it does at all. Curious to hear your thoughts on that.

Shadman Syed
I am frankly not an AI guy, but I would just tell you AI is a big thing and it is going to be a big thing in the future, okay. And it's still a big thing. So where I have seen issues, it is about, you know, it's going to, again, it's, there are certain models they're going to learn and they will tell you that, okay, whether it's automation, these are the things that you can do improve your automation. If there is any data that needs to be analyzed, they will analyze the data. If they need to do any AI models for predictive analysis, they will predict certain things for you, especially for monitoring, okay, that, you know, if you historically collecting data and based on that historical data, if they see certain trends, they'll tell you that.

The AI model will tell you that, hey, here are the places we think about maybe scaling up or scaling down or security detection, for a security team, that's a huge thing that they have security incident and event monitoring, that we look for certain trends in our field in cloud. This is where I see AI is going to help us. AI by itself is a huge topic, and different sectors, they probably are going to benefit from it in different ways. I still think that it may be a little bit early in the game, but for us, cloud infrastructure, helping us with automation, analyzing data, predictive analysis, security improvement, fraud detection, this is where I see that it will play into quite a bit and help us.

Matt Pacheco
Excellent. And then, yeah, AI machine learning has existed for a while already.

Shadman Syed
For a while. Yeah. Yes, you're right. And sometimes, five years back, I saw especially the ML stuff, machine learning stuff was already there. As I said, I've also worked on the IoT side quite a bit, and it's collecting data from these IoT devices and doing predictive analysis on that. Hey, this, based on the data that we're getting from the sensor, that there's certain, like, devices on factory floors, is telling us that, hey, it needs service. Okay. It was already out there. Probably that stuff is going to further improve with the new improvement in AI that we see.

Matt Pacheco
Yeah, very interesting. So you mentioned Iot a few times. I'm curious as to how Iot are we talking about Iot from like, a manufacturing perspective or a big business use perspective? Are we talking like, consumer level, Iot all across? I mean, it's everything.

Shadman Syed
Yes. I mean, it's. It's not just that. It just, if you look at automotive sector now, we have cars with. Basically, these are Iot devices. Okay. In our house, in my fridge right now, I have. It's on Wi Fi. It's collecting data. It sometimes scares me. I don't know what else is collecting and sending over that. And then, yes, factory floors, medical equipment. It is capturing data. That's where sort of like, edge computing also is coming into play. Where it's. We have now these edge devices, which is bringing computation, compute resources close to these edge devices, helping them do most of the computational number crunching work close to that, to the. To the source, and just, only transferring back to cloud just the main data that they need. Okay. I don't know if that answers your question, but if you have any specific questions.

Matt Pacheco
Yeah, yeah. Where I was, where I was getting was to edge computing and the intersection with cloud. Like, what are the benefits of leveraging the cloud for IoT devices and edge computing?

Shadman Syed
You can host number of products and services that can get the, you know, the data that you need, not all the data because the latency, that's just not possible. Okay. With edge devices now, even with right now, my major focus has been, and I should be saying this has been on Azure, Microsoft or Azure, they have their own edge devices, literally purpose built for working on factories. Okay? And you set them up, you manage them from cloud, it's collecting all the data there and it's just sending the data back to cloud where it's just depending on the application that you have, where it's just monitoring things, it's looking at that. If it needs maintenance or if somebody is there, then you can tell them how to make those changes, stuff like that. Yeah.

Matt Pacheco
Cool. And then overall to cloud strategy. So we talked about everything under the sun. The one thing I would like to talk a little bit more about is disaster recovery, business continuity. What role does that play in your cloud strategy? And it is quite critical business.

Shadman Syed
It is quite critical disaster recovery. You should always prepare for a disaster. Unfortunately, that's the nature of our business. Okay. Man made or natural. Okay. Depending on how critical your application is, depending like I work for a SaaS service provider, we have certain SLA's in doctor, we call them RTO recovery time objective and recovery point RPOs. If they are aggressive, your doctor strategy is going to be quite different and expensive. A little bit relaxed is going to be a little bit different. It can be. That's where I said that IAC comes into play. If you have a little bit relaxed disaster recovery, RTO's and rpos, then all you need is your production side up. Okay. You have the IC in place. Make sure that your backup, especially for your database, they're geographically redundant. Okay.

And in case of a doctor, you should have a clear playbook. How you come up with that claim playbook you do once a year at least an exercise where maybe you don't have to do an actual doctor for your production, but one of the non production environment, same application, set it up the same way as production and using IAC, provision your infrastructure there. Using your backups, just restore your backups for your database. If certain servers required, there's some static data that needs to be also restore them there. Just run your application, see if do your smoke test and everything. Everything looks okay. That's what your playbook is. That's a simple doctor. Now if you have more aggressive RTO's and rpos, then this solution is not going to work. Okay. Then you need to think about something else.

Either you have an active sort of a site that there's Georg as a service, both Azure and AWS also provide for it, or you can do it without the doctor as a service. You can literally have a sort of a regional or global load balancer where you have two different sites in different regions. An application works that way. The only tricky thing is database because you can only have one consistent data source. Just make sure your database is redundant. And in case of a doctor, all you're switching is just the source of your database. But the key thing is please, please do a doctor exercise once a year and also do some contingency planning. What that contingency planning is simple thing that your disaster recovery documentation mostly is stored in cloud in some wiki or confluence, and all that region also goes down.

You got to think about it where your documentation is. If you store it somewhere else locally, okay, think about people. I am based in east coast, okay? And something happens and I am responsible for doctor. If there's a disaster on east coast, there's someone central or west region can run this ER in real time. So think about also people too. So just make that as part of your, the whole business continuity and disaster recovery plan.

Matt Pacheco
Love it. Love the part about testing your doctor plan. That's a, that's a big one. Do it frequently. People who are listening and then also.

Shadman Syed
Both cloud service provider over here, they deprecate certain services and they make certain changes that they're in. The that way, if you do it every year to those changes, you capture those changes. Make sure that you have the most up to date playbook for disaster and.

Matt Pacheco
You can improve based on what you learn as well.

Shadman Syed
Exactly.

45:14 - Future of Cloud Computing and Final Thoughts

Matt Pacheco
Your runbook and all that. That's great advice. So we're getting to the end of the podcast. I have just a few more questions for you, some fun ones. If you can go back and restart your entire cloud journey, what would you do differently? Knowing what you know now?

Shadman Syed
I honestly would partner with someone. I did, frankly learn most of the stuff on my own. I'll find some sort of a mentor and a partner, okay. And this is what I also try to do that in my team. I'll have someone who sort of like would, based on my experience, learn from me. Okay. Nothing perfect. I mean, after whatever I have learned, it would be easier for someone to learn, for someone who has already gone through that journey so that I would do it differently. There's many things that I look back and say man, that was. I wish I had done differently, but there was no one to tell me that. So get a mentor, get a trainer, if that's possible. And then for the leaders out there, do it for your team. It's going to help them.

Not just help them, it's going to help you also, because if they do it the right way, it's going to give a good results for your team.

Matt Pacheco
That's good advice. Where do you see cloud computing heading in the next five to ten years?

Shadman Syed
Yeah, I'm not good at predicting stuff, but I guess I can make certain predictions. Yeah. I mean, it's a number of things that can happen. You already talked about AI. There's going to be some changes industry. What those would be, we'll find that out. The cloud native stuff is going to increase quite a bit. Initially started with just some databases, managed services. Now we have got this serverless. And even for DevOps, the native services, or for kubernetes or containers, managed services, stuff like that is going to grow. Edge computing is going to grow. And there's one thing that you, most of us have not heard is about quantum computing, okay? And that is a brand new idea, okay. It's still quite raw, okay.

And what that is that, you know, typical, like right now, the system, the computer system that we have is like we work on in bits, okay? Zero and n one. Quantum computing work on quantum mechanics and it uses quantum bits, qua bits, so you can have like five states, not just zero and one. So that by itself has a huge potential, especially for, you know, you do some crazy computational work. Okay? That is something to watch. Watch for. Okay. I, frankly, where I work, I don't see personally a huge need for it. But there are certain businesses out there, I would see that in next five to ten years, they might benefit hugely from quantum computing.

Matt Pacheco
Yeah, I would imagine as the amount of data grows too, I think every year the amount of data out there is growing, the need for quantum might be useful in a few years. I'd like to see where the progress is. I saw a news story about it, I think a few months ago that was in mainstream news talking about exactly what you're talking about.

Shadman Syed
It seemed like an edge idea, quite raw, quite new. Huge potential. Huge potential.

Matt Pacheco
Yeah, I agree. So, last question for you. What advice would you give to other technology leaders when approaching comprehensive cloud transformation initiatives?

Shadman Syed
We all have made mistakes in our life. Just learn from our mistakes. Okay? Partner with someone. Okay. Who has gone on this journey before. Okay? And this is what I would say that I have found, frankly, if there's one thing that someone can just take from this podcast is it's about your team, okay? If I make sure that you have the right team with you. If not, build the right team, okay. You should know the roles and responsibilities for team. Structure the team based on that. Hire the right people, okay? Or grow them. If you have the right team, sky's the limit for you. If you don't have the right team, no matter you work fortune 500, you have unlimited resources money wise. And you're right. Yourself cannot set up for success. So make sure that you have the right team with you. Okay?

Hire right person. Hire the right people. Make sure how you hire the right person. Know the roles and responsibilities for your team also. So make sure that you know what kind of team you need, okay? And lastly is try to get the respect for your team and the way you're going to get on the respect is in the moments of crisis. In our field, there are a lot of crisis moments. Lot of crisis moments. Just be there in those moments of crisis. That's how you're going to get your respect from your team. Not hiding behind someone or scapegoating someone. Make sure that in those moment of crisis, you were there with the team, you are helping them. And once you get that you have the right team and you have your respect, okay, you are set up for success.

Matt Pacheco
Excellent advice. Thank you. Shyavan, really appreciate you coming on the show today and sharing your insights and your advice and your expertise. It's been amazing talking to you and learning so much. Thanks for coming on.

Shadman Syed
Same here, Matt, pleasure talking to you. I hope you found this session to be useful. I mean, I really enjoyed it. Thank you. Thank you for having me.

Matt Pacheco
Thanks and to our listeners, thank you for listening. If you want to connect with Shadman, we'll link to him in the episode description. There's LinkedIn so you can ask them questions if you want, but follow cloud currents and tier pint on all your social networks. Get the podcast wherever you listen to podcasts and we will see you next time for more cloud conversations. Thank you.