The Future of Product Lifecycle Management

Tullio Siragusa (00:11):

Good day, everyone. Welcome back to Tech Leaders Unplugged. I am Tullio Siragusa your host, and I'm getting unplugged today with Andrey Knourenko, who is the VP of Engineering at Aras Corporation. Welcome to the show, Andrey. It's great to have you. Thank you. We're talking today about the future of product lifecycle management. Specifically, we're going to discuss the challenges of migrating fully customizable PLM systems to the cloud. We're going to dig into this in just a second, especially for those who may not know what PLM means. But before we do that, let's get to know Andrey a little bit. Tell us about how you got here and, why this is such an interesting conversation for you to have with us today.

Andrey Knourenko (00:59):

Well, I was born in the former Soviet Union, right? Get my master's degree in applied math and computer science over there. We worked for multiple companies in Russia and Israel, and the last 25 years in the United States. Somehow a lot of my work was related to enterprise software and kind of manufacturing things where it supplied, right? Even my first job after college was some kind of automation of some manufacturing processes back then. And here in the US I worked at the end of the nineties, worked for PTC. It's a large kind of CAD and pm company. And 2005 I joined Aras. I was a small company with big ideas on how to you know, transform PLM. And when I joined there were really 20 people, and we are kind of over 800 people right now, kind of one of the leading PLM companies. And what's more important is to continue to be a very challenging and interesting kind of, and this is why I stay there for so long.

Tullio Siragusa (02:15):

Great. Thank you. Andrey. For those who may not know, what does PLM stand for? What does that acronym mean?

Andrey Knourenko (02:22):

Well, PLM stands for Product Lifecycle Management. As you said on a fundamental level, this is a process of, you know, managing different stages of the life of the product, right? From ideation to design and then to manufacturing, and then to maintenance from a software standpoint, because obviously software is used for every stage of that lifecycle. It's a broad range of tools and applications, right? That allows you to manage those things from requirements to, I don’t know, a project to quality management, to design change management, etcetera, etcetera. So, the interesting thing is that although each manufacturer has some subset at least of those applications, right? Everybody does it very differently. The manufacturing process is very unique for every manufacturing company. And there is no out-of-the-box, right? You cannot just have some change management application. Everybody will be using manufacturing companies who want to have tools that allow them to adjust the software for their needs, right? To make it exactly hard to reflect the manufacturing process. And this is kind of where the challenge comes from. On top of it the complexity of products, right? In the last 20, or 30 years grew exponentially, right? So, I don't know, classical example, take your car 30 years ago, it's barely had any software. Right now. It's like tons of, in tons of microprocessors. And so, software drives everything, right? So plus, the company, of course, wants to change the manufacturing process and they evolve all the time, and they want their PLM system also to evolve with that, right? So, finally, again, the PLM vendors also release new, new software, right? To fix bugs, do upgrades, etcetera. So, moving an existing PLM system that might have gone through a very extensive implementation cycle to a new version of the platform is pretty expensive, right? And it's a very costly process, sometimes costing as much as a kind of initial implementation. And so this is what the company needs, they need flexible, powerful, and upgradeable systems. And this is where a comes with such a platform, right? We have such a platform that allows us to do that. And we, as I said, it was a small company. We are now one of the leading pillar companies. We have a large customer base. It's like Honda and Boeing and Microsoft and the list goes on and on and on.

Tullio Siragusa (05:32):

So, let's talk a little bit about the current state of moving to the cloud. Clearly, there are some benefits of moving to the cloud, more flexibility, accessibility, scalability, and dare say, also the ability to leverage, for example, advanced analytics and even AI capabilities, right? How's it being done today versus how it could be or should be done on, the cloud? What are, the complexities of making that shift specifically versus how it's done today, like as is versus how it should be? What are you seeing as a challenge around that?

Andrey Knourenko (06:11):

So, traditionally, PLM systems were deployed on-prem, right? Legacy M systems, it's just the products that the customers deploy in the data centers. And as an innovator, that's our product is work the same way. And we started looking obviously on a cloud a while ago. What's interesting there was very little interest in manufacturing companies moving to the cloud. And I believe the major concern, of course, was security, because it's definitely damaging for the company when, let's say your sales leads get stolen, right? But when your design documents of your latest model that you plan to release next year are stolen, right? It's a completely different impact, right? On your business, right? So, this is why kind of those companies like, we're reluctant to move to the cloud because the perception for the yes was that when we keep it close to our chest, right? It's more secure and things like this. I think over the last several years, that perception changed because of obviously increased requirements to, to the security and like the standards, it's very difficult to implement, costly to implement, right? By a company. Its, and cloud vendors also, providers now have much more tools, you know, to, to, to support those requirements. So finally, again, the cloud is now so you know, prevalent, right? It's like enterprise companies now kind of use a lot of SaaS applications, right? So why not P lab? So especially that's changed obviously with the covid, right? So, when everybody moved online, so in the last few years, we see increasingly increased demand and it's increased exponentially, right? For migration to the cloud. So, we start offering a few years ago our platform as a SaaS solution. And we have already several customers who do this and much more, in the pipeline.

Tullio Siragusa (08:33):

You said something interesting about the fear of the risk of tapping into intellectual property or even stealing intellectual property. Has, that been the primary driver for keeping, for example, PLM off the cloud? But that's been resolved for the most part in recent years. And I couldn't help but realize it's kind of like what's happening with the workforce, right? The mindset of everybody needs to be in the office. Yeah. It's the same mindset, right? You know, what, remote work is basically the same to a certain degree, the same challenge of having moved from on-prem to cloud applications. But anyway, that's a topic for another day. I was curious about some of the key benefits of making that migration. Would you say customization and integration complexities are, become less complex? Can you talk about that a little bit?

Andrey Knourenko (09:28):

So yeah, I think there are multiple kinds of aspects of that. Obviously, a company looking to reduce the expenses on it, right? And, but keep a high level of, you know, security, compliance and all those things to have kind of reliability of the systems, right? And to be able to do backup and all that's needed. And that obviously comes kind of easy with the cloud because that's the kind of job of cloud providers, right? The service providers take care of this.

Tullio Siragusa (10:04):

What about specific customizations? That might be built based on the machine type, right? Now you're moving it to the cloud. Do you have to rewrite it? What's the, what are some of the complexities?

Andrey Knourenko (10:15):

Yeah, this is a very good question because obviously the view cloud PLM system on the market before, right? But those systems typically have very limited flexibility, right? And typically, it's fine for maybe small and medium size company, but for large companies with complex products, with a lot of suppliers, et cetera, kind of, that definitely kind of doesn't work. So, this was a challenge because for a highly customizable system, right, like ours, to move to the cloud at least means that customers should be able to continue to do the customizations that we were able to do before, right? And if before, and we provided some tools around that, but customizations made on-prem deployment, it's kind of easy to manage, right? Because you can just, I don't know, copy files, right? To somewhere restart your IS and do things, right? And etcetera, etcetera. I mean, with the cloud, it becomes all, you need some tools, right? Like DevOps environments that allow you to do this. Keep in mind also that on a large customization project, usually, it's a team of solution engineers, right? That kind of work on this. So obviously we provide some low code kind of tools to customize, to do UI to do things, but ultimately in the end, it's all translated into some set of data configurations, files, and metadata, right? And if multiple people in the teamwork, work with that, it has to be somehow merged, right? And then deployed. So obviously there is some kind of source control system underneath. So, with moving the cloud, again, if you want to keep all this customization and flexibility, right? It should stay. So as a menu, you have to provide the first challenge, some DevOps environment that allows people to continue building those customizations, to test them, to deploy them, right? Even if it's kind of the instance of your dep, your deployment is running somewhere on the cloud and served by a service provider. So that's one challenge.

Tullio Siragusa (12:33):

I mean, just to, just to add to that IEM mean, I would think a lot of the concern is, well, we lose certain functionalities that are dependent on both the code and the machines that it's written on, move into the cloud. Does that have any impact on the business operations? So, it's not so easy to just move it to the cloud, you need some kind of a platform that allows for that conversion or translation. Is that what you guys are proposing that you provide a bridge to do? Can you?

Andrey Knourenko (13:03):

So, we do, as I said, provide customers with this DevOps environment that allows them to, again, continue doing what they did before our cloud deployments. Our customers' cloud deployments have exactly the same capabilities as OnPrem deployments. And this is quite unique, I believe at this stage, at least in the PLM domain because it's not very easy to do. So, in addition, if you wanted to run the cloud efficiently as a service provider, so you have to provide some cloud management platform, right? Because think about this every deployment, as I said, is quite different, right? The manufacturing process is different. So, each deployment for each customer although they all built on top of the same platform and might even share some shared services, right? But there is obviously they highly customize it. Essentially, it's quite a different result that like maybe one customer sees from another. Nevertheless, as a service provider, I am interested in somehow running this use as much as shared kind of resources as possible, right? To optimize my cloud cost, etcetera. So, I need a very efficient cloud management platform to be able to deploy those deployments, right? Those customizations for particular customers, monitor them, update them, kind of keep them up to date, etcetera, etcetera. So that's another piece kind of that's not very simple to build such a cloud management platform that allows you to do that.

Tullio Siragusa (14:48):

This is great. Let's dig in a little more on some of the concerns perhaps that organization might have in that migration. For example, maybe concerns around latency or you know, the user experience from a performance perspective based on network connectivity or system uptime. These are all things that don't come up as a challenge when it's on-prem, right? So, what's the sort of fix for those concerns?

Andrey Knourenko (15:20):

Well, it's definitely true. Although we have already at this point multiple cloud deployments for multiple customers, and they worked just great. I think obviously if the latency is, is a factor, right? And typically, when applications are deployed inside your local area networks, I mean, the latest is significant to lower, but I think there are a couple of points here. First of all, large companies, that's mostly our customers. They have very distributed by nature, right? They have offices in different countries, so doesn't matter if deployed in their data center, it's typically somewhere remote. It's not like the server installed the next room where the people sit, right? So that's one point. Another is I mean there are definitely tools how to achieve kind of low latency even with the cloud. But the third is definitely that's one of the things we, with migration to the cloud, have to work on a lot. We need to minimize kind of the traffic between, you know, the client and the server where the app, the server side, the application is running, right? So, this is required some changes may be in the application. And we kind of did this, and again, we have many at this point cloud-based deployments, and they all work.

Tullio Siragusa (17:02):

Do you find the need for planning, like things around change management from a user adoption point of view, as companies migrate from PLM to the cloud what have been some of the best practices you've noticed some of these multinationals deploy in terms of addressing concerns related to latency or addressing concerns related to functionalities that, that might need to be rewritten or might look slightly different as a result of that? How are organization planning for this kind of change and you know, in order to, to ensure good adoption as the transition happens, or some tips on how to plan for it? Yeah, that'd be great too.

Andrey Knourenko (17:48):

No, I think definitely yes, migration to the cloud is a big shift, right? For those companies. But I think that's the key. We try to make that migration as seamless, to our customers as possible, right? We try to provide the same level of flexibility in our product as in SaaS software and if it's deployed on-prem, we provide the tools that allow again, them to build those customizations, etcetera, etcetera. Obviously, there's some change in management inside the co companies right now. This is not the responsibility of maybe the company's IT right to, to run this application and keep that up to date, but the service provider although obviously it costs somewhat more because we now do the job of your it, but if you calculate all the expenses on the hardware, right? Keeping this hardware up to date, all this kind of need for all the security systems you need to implement to make that kind of secure, right? IT personnel, etcetera, etcetera, obviously running in the cloud is much more efficient. This is why we see such a large adoption mode of cloud solutions right now.

Tullio Siragusa (19:09):

You mentioned most of your clients are multinationals. I'm curious what is the market size for this? How many companies are still running PLM environments? You know, it seems the cloud has really evolved and matured over the years. Is there still a big percentage of companies that are struggling to move to the cloud or still dealing with having to plan all this, you know, network connectivity and, and, and IT staff and, and all the management that goes behind it? Obviously, you guys have built a business around it. I'm just curious about how big of a problem. Is this still a migration challenge?

Andrey Knourenko (19:49):

Well, first of all, you'll be surprised. A lot of companies, of course, it's a lot of them, maybe small medium, some companies still using Excel, you know, for kind of managing some of the processes related to the life cycle of their products, right? So, I would say Excel is the biggest competitor from the four pillars and vendors, right? So, but with definitely when companies implemented PLM systems it was a big hope, right? If you go back many years in general with the implementation of the PLM system, the hope was that, so now we have that great system where the single source of truth, all our information will be there, right? All our cycles of that life cycle will be connected, right? We will have all this beautiful digital thread where we can trace how the changes in, for example, requirements kind of affect the design and, and the manufacturing process and the other way around, right? In reality, unfortunately, it's a famous 10-1,000. So as a result of a lot of PLM implementations, only 10% of the functionality was implemented, and 10% of a user that they planned initially kind of have access to the system, but a hundred percent of the budget was spent, right? And obviously again, this is kind of, that's an unfortunately for a lot of companies that use kind of old legacy p m systems. So, I think this is one of the reasons ours was so successful. Again, we provide a much more flexible, much more kind of agile way to do this. And migration to the cloud is the next stage of this, because obviously now those moves to the cloud, right? Making a system kind of put into production will be faster because you eliminate that stage of setting up your own system, right? Higher than 80 personnel to kind of maintain it, right? To worry about all the things kind of to run physically in your data center, the systems you need, etcetera, etcetera. One important thing I, I think I mentioned with the thing that's upgrades, right? It's I mentioned this, but the upgrade is a big deal, and this is one of our big competitive advantages. I think right now when implementation is done after a few years, customers find that upgrades kind upgrade needs to be done, and then the upgrade is kind of very difficult. I think one of the challenges, again, in the cloud is to make sure that those upgrades can be done much faster and more automated way, right? We already do upgrades significantly faster than our competitors. There are some independent data that upgrades with are up to 20 times cheaper than with some of our competitors. But in the cloud, we see a lot of opportunities to make it even better, right? Because you can definitely, you can share some services between multiple customers where those services definitely that not hold the data or customizations that can breach the boundaries of the deployments, right? And that can be upgraded automatically. And then we introduce more mechanisms in the cloud that allow us to understand kind of the application, right? And maybe as you said, applying some that's our kind of plans to AI models, right? To understand this and generate some automated tests that can be applied immediately. The after upgrade is done, the customer has significantly less work and can certify the upgrade before it goes into production. So, there are a lot of advantages that companies have when they move to the cloud.

Tullio Siragusa (24:08):

Excellent. We're up on time, but I was curious as a final question how is AI coming into play? Does it help in as helping to assess what needs to happen, and the risks associated with dependency on local machines, for example, for moving functionalities and capability? Is that something in the roadmap for you guys, or is that part of how you do it? What are your thoughts on that?

Andrey Knourenko (24:34):

Well, it's definitely exciting new technology. We look at this starting from we now testing AI co-pilots right? In product development, right? So that allows to quickly generate at least some stop of the code, stop of automated tests right in they can, can be changed by developers. We see a significant potential increase in productivity for development teams on that. But then when comes to the customs customizations and deployments of our platform by customers, there are a lot of opportunities there, obviously from, again providing some services based on models that allow customers, for example, to make choices in design that are more environmentally friendly, for example, right? So, then the AI system will be able to gain, based on the data collected and model trained to suggest I kind of like, oh, instead of maybe this part, right? And assembly, we can use this part and requires significantly less.

Tullio Siragusa (25:55):

Yeah, brilliant.

Andrey Knourenko (25:55):

Yeah. And like this. So, it's a lot of ideas, definitely, that's an exciting new thing and we are looking at this and it's common.

Tullio Siragusa (26:04):

Excellent. Well, Andrey, it's been a pleasure to have you. We're over time. Thanks for being with me this morning. My birds here are chirping in, they're trying to get, be part of the show. So just stick with me as, as we go off the air in just a second, I want to announce what we've got coming up next week. We've got on Monday Ebenezer Schubert, who's the VP of engineering of OutSystems, and on Tuesday we're talking with Reza Raul, who's the CTO of Real Network. So, we'll announce the topics as we get closest or keep an eye out on your whatever, LinkedIn, I'm sorry, your social media platform. You're following us, we broadcast live on all of them. So, plenty of opportunities to catch us. Thanks for being with me. Thanks for watching this week everyone. We'll see you again next week. Have a great weekend and take care.