Speaker 1:
From the heart of Dubai, where tomorrow is being built today, to the world. Welcome to the CTO Show with Mehmet. Here, we redefine technology and reimagine possibilities. With Mehmet, delve into the riveting realms of AI, cybersecurity, and digital technology. Experience the thrilling highs and lows of startups. Immerse yourself in the spirit of entrepreneurship, and witness the future of business innovation being written in real time. Now, without further ado, let's tune in and explore the future.
Mehmet:
Hello, and welcome back to the episode of the CTO Show with Mehmet. Today I'm very pleased joining me, Ryan Pollyniak. Ryan, the way I love to do it is I don't ... I was telling you, I don't do these fancy, lengthy introductions, because I love my guests to introduce themselves. Tell us a little bit more about you and what you are up to currently?
Ryan Pollyniak:
Yeah, absolutely, Mehmet, thanks for having me on. So, Ryan Pollyniak, have been working in the IT business systems space for going on 20 years now, time flies. Started with ADP and the last 15 years or so it's been mostly in the Microsoft space, helping all types and sizes of businesses either begin their digital transformation journey, take the next steps, working through companies' growth and helping them strategize what's next from a systems' perspective. Of course, the latest and greatest, a lot of companies are focused on AI, a lot of companies are focused on cloud, so that tends to be where a lot of my day is spent right now.
Mehmet:
Yeah, of course we're going to discuss this in details now, Ryan, and this triggered the first question actually in my mind. You mentioned about the digital transformation and now AI came into the picture. Do you think organizations are ready to start this AI evolution in their businesses? Do you think that they have the foundation of the technology that they have established already, or there's still some work that have to be done before they can stop into implementing AI?
Ryan Pollyniak:
Yeah, I mean, that's a great question. And so, everybody seems to want AI right now, and the funniest description of that that I've heard is, "Are you looking, Mr. Executive, Mr. Business Owner to implement AI and to leverage all these great new capabilities in your business?" And the answer is, "Yes, we need it right now, but we're not sure exactly how or why or how we're going to get there." I think to answer your question, in short, most organizations are ready to start using AI on the periphery in certain ways, helping employees to be more efficient, not by replacing them, but by helping them to focus on value-added tasks as opposed to mundane day-to-day, non-value-added tasks, creating emails and recapping meetings and things like that.
Now, to get to the broader AI benefit case, I think a lot of companies have a little bit of ways to go, and typically that comes in the form of getting their data estate in order, because having data that's in disparate locations that's not properly structured and it doesn't make any sense or having data integrity issues, AI has to have the proper inputs if you're going to make any sense of it. If it's going to give you anything of value in return, you have to feed it data that is meaningful.
And so, AI is cutting edge, right? And the sky's the limit. I don't think anybody knows the tip of the iceberg yet in terms of how far AI is going to take us and how far AI is going to transform businesses. I advise our clients, dip your toe in, start using AI for the low-hanging fruit, so to speak, and then in preparation for what's to come get your data estate in order, get your systems in the cloud. Leveraging AI for on-premise systems is going to be from a technical standpoint, but more difficult have fewer and farther between use cases. So, migrating systems to the cloud and aggregating your data in a manageable and meaningful structure, I think those are the foundational things that businesses need to do to be prepared to take advantage of AI.
Mehmet:
Do you think, Ryan, to your point, the low-hanging fruit that I'm seeing organization are thinking about is just implementing kind of a ChatGPT on top of whatever system that they have? And you said this is just the top of the iceberg, right? Now, why do you think organizations need to, first, of course the data is very important, absolutely. But do you think also there are some misconceptions about AI in general that okay, it's just the chatbot part of the business? Don't you think also they are missing on reviewing the processes themselves? I mean, maybe they can add some automation also as well. Do you think that there's some misconceptions businesses they have currently, and how do you think these misconceptions should be addressed?
Ryan Pollyniak:
Yeah, so you nailed it in that there are some, when I say low hanging fruit, something like ChatGPT, or Grammarly, or some of the Copilot things that Microsoft offers in Excel and Outlook, these obviously can be used more tactically where you are composing an email, or where you're having something proofread, or where you're really generating content, avoiding manual data entry and creating more efficiencies. Note-taking is a great example. If you're on a meeting and you're recording it, having an AI assistant like you've got in Microsoft Teams where you can get a recap of that meeting and have the action items relayed back to you, which tells you, "Here's what you discussed, here's what you promised." These don't require an overall aggregation and cleansing of data. It's the more advanced stuff that requires that, right? If you want to have your accounting system go back and review your customer's payment history so that where most collection departments now, most accounts receivable departments now are relying on outdated methods to forecast cash flow as an example.
As you put things in the 30 and 60 and 90-day aging buckets and you rely on that and you spin out a report that says, "Okay, here's how many customers are going to pay me in 30 days. Here's how many are going to pay me in 60 days, in 90 days, right?" Well, that's kind of wishful thinking and hoping, rather than relying on a tried and true reality of what's happened previously.
An example of an AI that you would have to really have your data in order for would be, do you have a real customer payment history for your clients? Meaning, "X-customer typically pays 17 days early and Y-customer pays 23 days late." And once you have that powerful foundation in place from a data standpoint, then you can leverage some of the more powerful aspects of AI to go back and look and say, "Okay, your cash flow prediction is not just going to be based on aging buckets anymore, it's going to be based on reality over the last year or two years, however long. This is how your customers, your actual customers have actually paid. Here, Mr. Owner, Mr. CFO, Mrs. COO, here's what you're going to expect in terms of cashflow." And it's far more realistic. That kind of data crunching is a different story than using a ChatGPT to compose an email, for instance. But there are certainly some, that's more where the low hanging fruit is on the latter part of that.
Mehmet:
Absolutely. Now, back to something you mentioned also in the introduction about the data. And they sit everywhere. The data sits everywhere. Now, why do you think until now, and we are in 2024 and we've been collecting data for such long time. And I know this for a fact, I used to work in the data management space for some time also as well. People they don't know until now where their data resides, they don't know what kind of data they have. It's kind of now it's a common word we started to use a lot. I mean, it's a word used before. I mean, the recent episode we repeated a lot. It's kind of a technical depth when it comes to data and how organizations are able to organize this data.
Saying that, and knowing that this is a challenge. How do you advise, Ryan, these organizations when they should start actually, from which point they start? Of course, the structured data, maybe it's easy. Anything which is sitting in the databases, it's there, they know. But they have, as to your point, a lot of things, few in the cloud, some in the edge, some other places we don't know where. In order to your point also to get this data so they can build on top the AI system they aim to, how they start, from where they start. And what are some of the, I would say, steps usually you advise them to do to get things in order?
Ryan Pollyniak:
Sure. Yeah, that's a great question. Primarily I've lived in the ERP and CRM space for the last 20 years. And that's a great place to start if your organization is outgrowing the systems that maybe you started your business with, or maybe your organization is taking the next steps into moving from a small to medium-sized business market into more of the enterprise market and you're looking to make that leap to the next evolution of systems. Aggregating as much as you can into one system without sacrificing the quality of any particular module is very important. And so what I mean by that is, we see all the time a company will have had an HR director go out and get an expense management system. And they will have had a sales director go out and get a CRM system. And they might have somebody in the shop floor running the manufacturing department go out and find a manufacturing NES system, manufacturing execution system or a warehouse management system.
Empowering people in each department to make system-based decisions on their own is going to start to have consequences down the road. And the reason for that is, when you go to make a big core fundamental system change like ERP, which really is at the hub of everything in your business as it grows, having to rip out some of those processes and some of those systems where those decisions have been made in silos is going to cause business disruption. It's going to have a tremendous amount of cost. Forward-thinking and planning out, okay, here are the areas of our business where we want to capture data and where we want to automate processes and where we want to move to the cloud or move out of Excel where a lot of people manage a lot of tasks these days.
And that's a great thing to ask yourself as a startup business owner or as your business grows, "What are we doing in Excel? What are people doing on notepads and managing outside of a core business system?" And the more of that that you can bring into a true cloud-based software-as-a-service system where that data will be as long as it's properly implemented. And that's another part of it, don't go out and try to figure it out on your own. Work with professional to implement the system properly and it will pay massive dividends down the road. But having as much data aggregated that you can into a system that maybe has some AI capabilities, and there are lots of them out there. Of course, I'm mostly familiar with the Microsoft stack, but a lot of them have AI. And then leveraging, having your applications in one system, not only does it pay dividends in terms of being able to then leverage something like AI to make sense of that data. But it reduces integrations, it streamlines process automation, it pays dividends in a number of ways.
Having siloed systems is a big pitfall, I think, and especially as companies grow, just going and picking something off the shelf. Because you need what's point solution for that particular requirement will lead to disparate systems and integrations and a lack of consolidated data. I think that would be one of my recommendations to get started.
Mehmet:
Absolutely. A couple of things, siloed systems and what we can call a shadow IT is something, because someone decided to your point that there's a good procurement system, I'm going to go and buy it. And then they start to, without getting back to the IT. I used to live in that world back in time. Now, maybe this is for startups are lucky enough, Ryan, because now they can decide. Or even it's still, if they are in the scale-up mode, they can still decide and they have the flexibility going with these modern systems. Now, sometimes we have these big companies which have been in business for quite some time now and they are using some legacy systems. Of course, these systems probably they used to be top technology when they acquired them, but now they are outdated. And honestly speaking, on many occasions when I used to see these, either when I say legacy, either the system itself is a legacy system, I mean from architecture perspective, or sometimes even it's obsolete.
But the reason they cannot go out of it, they come out with a lot of reasons, "Oh yeah, we've been using it for all the time, but we cannot do this." Now I'm talking about the legacy systems. And first I want to hear your opinion, why actually it's a problem to stay on legacy systems? And what would be some of the steps you recommend for successful migration to the cloud? Now of course, and before you answer, Ryan, I know some people say, "Hey, I'm regulated, I cannot go to the cloud. This is another discussion." But I mean, in general we're talking here about majority of the enterprise that they can actually migrate to the cloud.
Ryan Pollyniak:
Sure, yeah. Legacy systems, and I've seen in my couple decades doing this now all different situations. Let's start with the basic one, right? If you're on a system that's unsupported, and this system is absolutely business critical for your business. And maybe you've got someone in the back who can keep it running for you, or maybe you've got a third-party company who's able to support it for you, but if the OEM, the developer of that application no longer supports it, it is time to look for the next iteration of what you're going to do. And that may be painful, and that leads into another discussion of how to properly implement going forward, which we could save for another time. But if you've got a lot of customizations and a lot of changes to the system that have been made to accommodate your business requirements, it may be painful, but it's not as painful as having a business critical system that is unsupported going down and grinding your business to a halt. That's worst case scenario.
Ransomware comes to mind. Because outdated systems lack security patches. And on-premise systems, no matter how good the company or the internal IT team is that manages your firewall, are ripe for a cyber attack. And this is where we see all of the ransomware, all of the cyber attacks happen. It's on-premise, it's something secured internally. Now, outsourcing that to a company who is providing a true SaaS solution, who has huge R&D budgets to keep these systems safe and has a documented track record of doing so, Microsoft is great at this. There are other companies that are great at it. It used to be scary to put your data in the cloud. And you mentioned regulation. Now, there are some entities that we run across where it's either an internal decision by executive leadership or some kind of regulatory requirement to really not go to the public cloud.
There are typically options for that. I mean, we see the government cloud that is ultra secure and compliant. The U.S.'s Department of Defense has systems in the cloud. There are all kinds of very highly regulated industries that have cloud deployments. I think the biggest hurdle there is sometimes getting leadership to understand that in all reality, you're much more secure now in a true SaaS environment secured by one of the Microsofts of the world who just locks this stuff down, closes off the backend, and you don't see ransomware attacks there, you don't see cyber attacks there.
The other advantage of getting to a true software-as-a-service offering, and just to pause on that, software-as-a-service I'm talking about something multi-tenant hosted by a provider. Not just putting your server in another location and paying someone to manage it for you, that's more of your managed IT. And that can have some advantages, but getting to a true software-as-a-service application provides the advantage of getting constant updates. Typically, several times a year you're going to get bug fixes, you're going to get security patches, you're going to be updated to the new version so that you don't land in that same spot that you're in on this old outdated system that you don't want to touch for fear of cost and business disruption. Getting to the cloud helps really with all of that. I would say somebody on a legacy system that is unsupported, that is on-premises, consider the business risk as risk mitigation as really one of the leading reasons to migrate.
Now, of course, all of the benefits of gaining efficiencies and leveraging bells and whistles like AI, which are emerging are great. But first, get yourself on a solid foundation. We've seen situations with on-premises outdated applications where big companies, I mean, several hundred million dollars a year in revenue companies go down, can't ship, can't invoice, can't pay their vendors. And if you imagine the dollar cost of that per day, it dwarfs anything that you would have to endure to get your systems updated. It's like a trip to the dentist, right? If the tooth is hurting, go get it fixed. It's got to happen. It's absolutely critical.
Mehmet:
Absolutely, Ryan. And I remember myself one time I've seen, and this was yeah, 2020 I think, before the pandemic, beginning of 2020. I've seen a system running Microsoft SQL 2005. And I asked the client at that time, I said, "What you would do if something goes wrong?" He said, "I don't know. I have no idea." I said, "Why you don't change?" He said, "Because whoever built the application to us, he designed a way that you cannot upgrade the database. I cannot remember the reason." But I said, "You're running in a single point of fail. You better go and find another software company that developed that, and probably the cost will be much less than if you lose the server and you cannot recover." Because I remember I was working in the data protection space and we could not back up that server because this was running Windows 2003 and SQL 2005 and said, "This is a disaster." To your point, Ryan, I remember seeing a lot of news at some stage, because some companies went out of business because their systems were legacy, and when cyber-attack or whatever similar disaster happen, you cannot recover, so you're out of business.
And to your point also as well, and just to make it relevant also from geography perspective, the hyperscalers, all of that, including Microsoft, they've invested in a lot of regions. Including where I live here in Dubai, in Greece. We have a lot of regions now, there's no point about data sovereignty because the data stays here. Absolutely, 100% I agree with you on this. But yeah, hopefully people will understand the need to really modernize. Best case, at least modernize on-prem and then try to shift to the cloud. Because I don't know if you agree with me, Ryan, there is a trick here. Just shift and lift doesn't work all the time, right?
Ryan Pollyniak:
Nope, nope, you're right. And there are a couple of reasons for that. Number one, there are technical limitations. If you're going to go, especially if you're changing platforms, one ERP to the next. And one thing that I always find myself explaining to executives from business owners is that, no, you cannot bring, if you're going to go from X, Y, or Z ERP to a different one, you can't bring all of your detailed transactional history typically over. Because the data structures are different. And all the underlying interconnected payments attached to invoices and item ledgers and everything else, you can't shoehorn it into the new system. Now, that said, I follow that up with, nobody has ever said, "Okay, we don't need our old data. We'll just start fresh." Of course you need it, but that can be accommodated by a data strategy that involves archiving old data into a data warehouse.
Cloud based, absolutely. I would recommend, especially if you're going doing this for the first time. Aggregate that data, pay somebody in the data consulting space to help you create a unified reporting model. Because you don't want silos of data again. You don't want to take all of your data from your legacy system and have it in one silo and then all of your data in your new system in a different silo. You need context. You need trend analysis year over year. And so for that to happen, aggregating data into a data warehouse, creating a unified data model where your new system is updating that data warehouse in context with your legacy system data, this is how we approach this almost all the time now. And so, no, you can't always bring your data to the new system, but there are modern ways to handle that.
And if you're a younger company, maybe you're a startup and a data warehouse is overkill, leverage something like Power BI or there are other data visualization tools out there where you can actually aggregate data in the cloud service itself. But it has limitations as you grow and then maybe you migrate to a more enterprise level data architecture at some point. Lots of options there. So, data is one side of it. The other side of lift and shift, there's no value in it typically, or very limited value. Because you probably implemented the legacy system 10 years ago, 15 years ago, even if it's five or six years ago. Your business has likely evolved, the requirements is likely evolved. And the business system functionality for the target application is going to be much different than what you have now. You have to go through an evaluation to match up those business needs with how the solution is meant to work out of the box.
And drive
