00:00
Welcome everyone. Um I know lunch runs from 11 30 to 1 30 our session starts at one, it was a way to allow people to be flexible in the middle of the day. So all sessions uh there may be people trickling in and that's OK. Um So uh with that, um the next session you're in the port works track.
00:20
And let me uh put up the before I put up the first slide that this session is really going to talk very much about platform engineering, right? For those of you who've been with us the last couple of days in different discussions, you know, that the big trend is about how companies are putting together platform engineering teams for, you know, velocity innovation and so forth,
00:44
right? So with that in mind, um what we thought we'll do is have a customer come up first and who's actually in the first phases of that journey. So it'll be a good way to start the discussion. Um And then get into the 2nd, 2nd part of the discussion, we'll go deeper into platform engineering. What are the things you need to consider,
01:04
particularly if you're an infrastructure um you know uh folk, then you want to think about what, what kinds of things you would be doing. So before I go there, show of hands anybody here other than Kendra who is already a Port works customer, you are a port works customer and um rest of few um pure customers using flash Ara flash blade, things like that.
01:26
Yeah, fantastic. So with that, let me jump to introducing Kendra. OK. Um Heritage Solutions is a customer of both pr as well as Port works. And Kendra mccormick is been gracious enough to come in and talk about really, she's gonna talk about their journey, Kendra. Can you come up, please?
01:54
Um And we'll just talk around here. We're just, you know, making it impromptu. Um The I wanted to first introduce the company. Um And so people have a context for what it is, you do and what your size of the problem is and therefore they can associate them. So I got to get the clicker.
02:14
So if you don't mind, sorry about that. Um So it's clear as to whether they can make a connect in their minds about what it is they are going through and whether uh your experience is very useful for them. So thank you very much for joining us here. Of course. Yeah, absolutely.
02:30
So I can tell you a little bit about health edge, health edge is both a software development and a hosting provider. We own and develop our own health care software. Um Our software, at least my division of the business is focused on claims adjudication rep pricing. So what that means is when you, um, go to the doctor have a procedure, fill a prescription um, in that pharmacy or the doctor's office,
02:56
somebody has to check your health plan and what's covered, what's not covered, um, what your co pay may be and then that goes back to your health insurance company and they do the same cross checks, they may pull out something that doesn't process properly, um et cetera. And so that health insurance company, the hospital where you may have a procedure. Those are our customers that are using our
03:19
software to perform those functions. Um It's really important to you as consumers and patients, right? Because you want to make sure that your coverage is available when you need it and it's priced properly. So you're not paying more money out of pocket than you should be.
03:34
And when you have questions about your coverage, you want to have a clear layman's term answer um to understand what your coverage may be and what your financial obligations may be. So health edges software, I particularly work on um Health rules payer is my sector of the business. Um and I manage a team of eight engineers onshore and about 10 engineers offshore in Pune, India.
04:00
And my team is responsible for everything from the operating system back to the physical data center, so operating system including security updates, um all of that stuff, um the infrastructure, the servers, the networking, um all the way back to the two physical data centers that we have, one in Massachusetts and one in Ohio. Um And you know, with the understanding of what our software does,
04:25
um I'm sure you can understand that up time is really, really important because if you go to the doctor or you go to the pharmacy and our software is down for our customers, that means that you may not be able to get the services or the prescriptions that you need. So 24 7 business, um our companies are required to meet compliance with Medicaid and Medicare, which are very, very strict, full HIPA compliance.
04:51
So, you know, security is really important and we need to be able to do that fast, right? And so health care is a lot of data. My business in particular is committed to making health care data available to patients um forever. So, um that's a lot of data that continues to grow. We continue to gather more and more metrics
05:11
about our patients and that needs to grow. So we need a place to put that and we need a place that can process that and make it accessible at a very quick rate. Um But at the same time, it's important to us to reduce the cost of our operations so that we can pass that, that cost savings over to the patients, right? You don't want your health care costs to go up. And a piece of that is your health insurance
05:34
company and a piece of that is your health insurance company using a software like like health hr payer um to be able to process claims very quickly with as little human interaction as possible. Yeah. So the thing when I was reading about health a a little bit more to understand their business, you know, in, in North America or particularly in the US, if you think about the number of health care
05:58
companies that are there that provide these sorts of insurance programs and coverage for patients, you know, 120 customers is 120 health care customers is a sizable deal. And so it seems like, you know, close to 1/4 of the US population is covered by, you know, health solutions for um health care customers. So that's really an important company and has been around for quite a few years with a few
06:26
acquisitions along the way um to put together these kinds of software solutions for them. So um with that, um Kendra if, if you don't mind describing your kind of journey um and you know your applications, your solutions and where you started your journey for containerization and kind of why. Um and then we can talk about what the results were with that.
06:49
Yeah, absolutely. So I've been at health edge for almost five years. Um At health edge, we like to joke, that's like a lifetime um because we do so much, we do it so fast and so much change has occurred for health edge in the last 3 to 4 years. Um Even during, during COVID, um when I joined health care,
07:05
we were technically a start up. We had 325 employees today. Um Health Edge has acquired three other businesses. We are now about 3000 employees and have gone from about 25 customers to over 100 customers across those business units. Um So it's been a a very quick journey in the last three years.
07:26
Um All those acquisitions happened during COVID. Um And so, you know, the whole going back to the office debate was not a debate for us. We don't have enough chairs for butts, right? Because we've expanded so much. But that expansion um has happened so quickly that we need our technology to be able to scale as quickly as our business has.
07:45
But at the same time, um we don't want the, the, the financial commitment to scale at the same rate, right? We want that to scale slower. Um So we have had to very quickly turn over our technology. Um We're still very much in the process. Um but we're going from a hyper converged all flash, but hyper converged nodes to our pure arrays with Lenovo nodes for compute.
08:10
So with hyper converged, if you're not familiar, you basically buy a big rack and it includes all your compute, all your storage, all your networking, your heating, your cooling, all of that one big node. Well, most of those, especially the platform we're on, that's about 100 and $50,000. Well, what if you just want to expand your storage?
08:29
You have to pay 100 and $50,000 and expand your storage, your, your networking, your compute, et cetera. So that very quickly, quickly drew us down the pure line because we were already all flash. So we wanted to stay all flash. Um That was not up for debate, but we needed something that we could scale our compute at a different scale or a different
08:50
pace that we scale our storage. Um So we still have some of that hyper conversion. And actually, I was, I was telling him that during lunch today, I found out that we finally completed the P OS to double our pr footprint to four petabytes as a service where ever green one customer. Um And what that means for me is we're pulling that hyper converged out of my data center
09:15
completely. We're going all orange, right? And this is a big deal for, for, for my team because we are trying to explore new paths like cloud computing in the public cloud. We're in an Azure space. Um and as well as containerization um with containerization. You know much of what we do like our software, the software itself is the same for all of our
09:38
50 hr P customers where it's different for each customer is their database, right? The, the individual data that's available for that customer. And because of HIPA compliance where we, we use network segmentation to maintain separation between those customers database. But does that mean that we have to have 50 instances of our software that looks the same
09:59
right now? It does because of that segmentation. And that's where we come to containerization. If we can containerize HR P, we can have a handful of instances of HR P that then reach out to the customers databases and still maintain that that network segmentation, right? So um what we're talking about is not reducing headcount but limiting the need to expand
10:23
headcount, right? Which then brings down our operating costs, which then we pass off to our health care customers. Um And it allows us to scale faster because we really just need to look at scaling the database, not the database and a number of attached third party applications as well. Um And it allows us to um speed up our upgrades um reduce downtime because with
10:46
the containerization, you can do real time upgrades. And because we upgrade our software, typically, customers get an upgrade from our, our, our software, either a maintenance release or a version release once a month typically. Um And so with containerization, we hope to speed that up and make that more resilient for the customers and hopefully almost zero downtime.
11:08
So the scenario before um you went to the container, was this the information based on, you know, 100 total DB instances, 50 customers, um you know, every month because there is an upgrade process, you were going through a planned outage. Um you know, so there are instances of outages and so forth. So since then with containerization I wanted,
11:31
you know, have you talk through the the experience and what that meant in terms of real impact for the business. Yeah, it's it's been been awesome. So the the data he's showing here is particularly one third party application that we use. So if you think about all of the information that's in your health insurance software, a lot of that requires addresses,
11:51
right? Because taxes are different by your address and regional costs are different. So every patient may have at least one street address, every provider has at least one street address, every hospital, every health insurance, there's a lot of street address data and we need to make sure that is accurate. So there's a third party software called
12:11
Melissa Data, which basically provides us correct address data with monthly upgrades to make sure that we have all the correct legal addresses. Traditionally before port works with pure, we had an instance of Melissa Data every single one of our HR P customers. So that's 50 instances. That's an application server.
12:29
Um That also includes typically two, sometimes three copies of that database. And that database is large because it's every street address in the, in the continental United States. And we had to do that 50 times. So that's a lot of storage, right? And that's where we're saying at least 50
12:45
database instances. That's just for this third party tool. Um And it's one of many third party tools. So, and that database gets updated once a month to make sure we have all the current stuff. So we have an engineer, an application engineer that would sit down every month and go to each individual instance for each customer and upgrade that database. And what that meant was pulling down another
13:08
copy of the new database with the two old copies there just in case applying the new database and saving a second copy of that database in case, you know, there was an issue with it and doing that 50 times every single month, it's a lot of man hours. It's a lot of I call it monkey work, you know, it's not skilled work, but we have to have skilled engineers in there.
13:30
Um And what we've been able to do through containerizing Melissa data is we have one instance of the Melissa data application and one maybe two instances of that database because again, that database looks exactly the same for all 50 customers. Why do we need 50 copies of it. Well, we don't with containerization. So port works with pure has allowed us to reduce that.
13:53
So that's one monthly upgrade on a single instance of the database and the application versus 50 previously. So that was typically about 400 man hours a month to do that upgrade is now down to four hours a month and it's starting to shrink because we're able to do it even faster. That's fantastic. And you're just beginning the journey right now.
14:14
So that's one third party application. We haven't even done that with HR P yet. That's right. So Kendra, I know that you're starting the journey. We want to be a part of the journey for, you know, all of your, you know, and help you all through the process. But I think that's eliminating for the group here to understand how,
14:30
you know, like mentioned in the previous presentation, which uh which you weren't at that. You always start small, you get your wins and then build from there. So that's a good example of it. And I wanna scream that from the rooftops. Um I think that's difficult for business leadership to understand.
14:46
Um We in particular, in the health care space have gotten a lot of pressure from our customers. Like why aren't you guys in the cloud? Our customers don't even really know what that means. They just think it's the next best thing and we're laggards if we're not in the cloud So my, my leadership is like, let's get to the cloud, let's get to the cloud.
15:01
And I'm like, hey, now our application is CPU heavy. It's compute heavy, it's storage heavy. What's the most expensive thing in the cloud storage? And CPU? Right. So does it make sense to push a hr P our, our, our main application to the cloud when we also don't internally, we haven't grown processes and skill sets around the cloud.
15:22
So we're still learning. So does that make sense? I personally don't think so. I think like you said, take the small wins. Let's look at third party tools that we're currently doing that are very consistent, low maintenance, uh low resource. Let's put those in the cloud. Let's see what that looks like.
15:37
Let's let's see how that improves other business process. And let's allow our engineers to learn from that experience before we start putting the big guns out there, right? And I think what we're learning from that is that maybe it doesn't make sense for hr P to go to the cloud, maybe it makes sense to make put our third parties there.
15:55
Um And we're going to have some customers that never want to be in the cloud. We're talking health care data, right? A lot of people don't necessarily trust the cloud with their personal health care data. So we're always gonna have to be, I think partially in the data center, even if we get hr P in the cloud. Eventually, I think we'll always be in a high
16:12
hybrid place. Um And again, that's where works and pure has helped us because nothing we do with pure will limit us from putting some workloads in the cloud or putting them there and pulling them back if we want to, if we decide it's not a right fit. Um It gives us that flexibility to decide what's best for our business and start with
16:32
those small wins. Um and, and learn from the experience. Fantastic. Thank you so much, Kendra. I appreciate it greatly. Thank you. Thank you. Thank you guys with that. I want to switch the topic to um platform engineering and Kendra mentioned a few things
16:51
in there in terms of portability and so forth. We'll dig much deeper into that. I want to introduce Ramari who is the head of uh platform uh products and engineering for the company with that. Alright, Kendra. That was fantastic. You know, it's II I just like to see the journey and how you were able to optimize your
17:10
deployment and realize the savings. Uh And you know, kind of like take your entire company forward. That's inspiring. And you know, that's why I'm so excited to kind of work in this space and work with our customers because our customers constantly are transforming their organizations and you like the challenge you mentioned,
17:29
I could totally relate to. It is like everybody wants to jump right into cloud or containerize everything and like, oh, you've not done that. But yes, there's a lot of practical difficulties so that crawl, walk and run and start small and building it out. Absolutely the right strategy. So, so excited to have you here.
17:46
Uh, you know, this is, uh, uh, lovely to have all of you here. Uh, uh, it's, you know, I wish we had more folks, but, you know, this is, this is great, you know, like a focused audience than like, you know, like a full, right? So I thought, you know, what I'll do is uh of course, we have a couple few slides.
18:03
Uh but since we're in Vegas, maybe it's, it's, is it probably a little bit uh good to play, right? It's like this case, you're all winners because you're already here, right? So a couple of questions, let's see who has the right answer now. Port works. Team. No brownie points for you for answering those.
18:23
All right. So what do you think is the most important commodity in a modern business today? Why? Two winners? But you're the customer? So you win, right? Data. Absolutely. It's a modern com.
18:42
It, it is the most important commodity. And of course, you know, like you're a genius, sir. Data is the most important commodity in the modern business. So what surfaces the next question? How can businesses, what do businesses have to do to get the most out of their data?
19:00
I mean, what do they have to build, to get the most out of their data applications? Howard? Yes. This time he won. I think I see a duel building here. Yes. So you now the enterprises and you know, businesses have to build applications like what health they just done for your customers,
19:23
right? You spoke about having all the street addresses, a whole bunch of data, but you aggregate all of them together, kind of build the benefits plan, but you know, you offer it to your customers, you know, and so businesses have to take the data, either collect data, you know, get data from other sources, harness it, build applications and transform that into
19:41
a business value, right? So of course, you had to go build applications. So if if you have the most valuable commodity is data and the most valuable effort is applications, who's the most valuable resource. Uh You know that the company has to invest in to build that. I mean, not that you know, marketing is great. Absolutely.
20:02
Sales is extremely important but yes, the most expensive resources are engineers or developers because that's where you know, we end up paying a lot and we want output, right? So you have to in a modern enterprise that values data built great applications has to invest in their developers and engineers and you know, and bring the best and the brightest and most often you compete for them.
20:28
So you pay a lot and you don't want those highly paid highly valuable resources that developer time to be wasted, right? You want them to be the most productive. So driving developer productivity and enabling the developers to be successful and being able to create high value applications is a real game, right? Because in the modern environment, in the modern enterprise,
20:53
if your business is not innovating, constantly, not driving new innovations out into the market, not you know, enhancing their user experience, you're going to be left behind. So developers and their time and how they can constantly create and deploy new apps and how to, how we can save them time becomes the most important thing that we need to solve in order for all the businesses to be competitive.
21:17
But where do the developers spend a lot of time on? Of course, there's a lot of water cooler dots. Absolutely. I do spend a lot of time there myself, right? You know, and then, you know, like obviously like learning new stuff and all of that. Um you know, brainstorming, you know, of course, the management kills developers,
21:37
a lot of meetings, staff meetings, this meeting, that meeting, right? Take all of that out. Those are all smaller. The most development teams spend very little time there, right? Those are, you know, the most time they actually end up spending is waiting for something, they're constantly waiting for
21:52
something, they're waiting for some VM to be spun up, they're waiting for some licenses to be available, they're waiting for some, you know, uh uh access to some uh uh software that they don't have access for. So they are constantly waiting for something like any system you have wait times it becomes less efficient, right?
22:09
So you know what, how do we help developers? So developers need to have self service, right? If developers can achieve, get most of the stuff done to run their applications, to build their applications, run them in production, constantly update them, then they can be a lot better, a lot efficient.
22:33
But can they really have self service today? Do they have like, you know, our infrastructure and everything we have today in place, all the software is it self service ready? Right. Not exactly right. I mean, if you go, you know, typical, you know, infrastructure today like a little bit,
22:47
you know, like you can go the legacy architectures BM you know, cloud instances, you know, there's, there's a lot of a lot of that is not self service ready because they are very machine centric. You know, you could go to an environment where you have virtual machines or, or servers or cloud instances. They are very machine centric and a lot of
23:04
developers don't have control over all of the all of that infrastructure, but they would like to get more control over the infrastructure. They like to declare how to move that infrastructure along to their needs. So that their applications can be built and run and managed at scale by themselves, right? They can easily update the user experience,
23:23
they can easily deliver a better user experience a lot faster if they have control over that infrastructure, if they can decoratively make that infrastructure work for them. So that is where you know, cities and containers come in cities, you know, delivers a platform on which developers can build and run their applications and can decoratively
23:48
work with the infrastructure now. But that's not just, you know, that's not, that's not just sufficient, you know, just to we just have right? Because you need a complete platform, right? You need like everything like you know, just not just the orchestrator, you need the uh you know uh the networking to go with it,
24:05
you need the security to go with it, you need the storage platform to go with it. So you know, all of this needs to be delivered to the developers so that they can bring their applications and then they can run it on their own. And that is where platform engineering comes in, right? What we see as like the Devops teams and you know, the infrastructure teams coming together
24:28
is how you know all of this infrastructure comes together as a platform, right? But you know, that means you have thousands of developers in an organization and then you have very few people in the platform team supporting them, right? They're almost constantly overwhelmed, you know, we have customers uh who, when they started out,
24:49
they had thousands of developers and they were like 15 people in the platform team and they never took uh a break during holidays, there were weekends, they were burning time and you know, they were trying to solve the problem of how do we offer self service to developers but not having to tinker with the infrastructure of the applications constantly, right?
25:09
So if you, you know in the modern enterprise, you see this is a real problem where as more and more uh apps move on to containers, move on to self service, you empower developers more. There's a lot that gets put on the platform teams. How do you scale them from DEV to test to production?
25:28
How do you go like, you know, yes there is, you know, of course, we have a most valuable commodity to build apps, make developers effective, build a platform and help them kind of give them a self service platform, they can build and run their apps. But how do you scale it? But how do you make it work?
25:44
That is the difficult part, you know, like it's it's an interesting challenge but you don't build a solution looking for a problem, right? And you want to build the right solution, right? So that the solution doesn't become a problem, right? So how do you how do you build something that scales and how do you run this uh so that, you know, your business can take advantage of it.
26:03
And that's where you know, platform engineering teams, you know, come in and then they have to go build this cloud native platform that continues to deliver uh you know, applications and drives more efficiency. So what does a cloud native platform attribute? What does it have? Right, what is, how do you what are the, you know, key attributes of a cloud native platform?
26:26
You can see that you know it needs to be, it needs to go well with continuous app deployment and integration and you know all of the C I CD and everything needs to be built in. You need to be able to build applications and run applications anywhere. You have to be all the you know all the applications you build, you know when you're building a cloud native platform, you need to be able to build an app,
26:47
run it anywhere, be easily able to replicate it in any other infrastructure like you mentioned, right? Like you have your applications in the data center, you like to be able to tier to the cloud, you'd like to be able to run a few instances within the cloud. So when you're actually building a cloud native application platform, you want to make sure that you build stem cells,
27:06
you build application stacks that can be easily replicated everywhere. So you can go to where your customers are. We are customers who run their applications in the data center, but they also have their instances running in a few clouds because that is what it is closer to their customers, right. So you need to be able to go and make your applications portable and make it like, you know, neutral uh to any,
27:27
any cloud. So, and that is, that is something that is kind of like that makes you more agile, you know, you know, makes makes your apps closer to where the customers are. The other thing is V MS are not going away, but you know, there are going to be certain apps that are going to be in V MS because most of the time
27:46
the developers have gone or there are some specific requirements that for that app, for example, you know IP ranges or DNS names cannot be changed. So those apps will continue to be in V MS. But many of the modern apps will be containerized and will run in containers. But a plat this platform has to unify both V MS and containers and be able to kind of
28:07
orchestrate it together and do all of this, give a global name space, give a global view but be completely secure and be multi tenant. Now, if you build that, if you have that, then now you can realize the vision of taking the data building faster applications, delivering radius experiences all through self service. And your entire development pipeline is a lot more efficient than it was ever before,
28:36
right? And that is where the cortex platform comes in, right? Why how it is is for data, you know, it's for applications port works is for data, right? So port works is no more than just a storage engine. Of course, it has storage in it, but it is more of a data data orchestrated data platform,
28:56
right? The port platform consists of core storage services. You're all familiar with port Enterprise, which is our flagship product. I think a lot of customers have that in production today. So port enterprise is a storage product, does all the storage management, volume management and everything.
29:11
That's part of the port platform. We also have port backup that does data protection and business continuity in Dr for uh backup services. And we have the B CD R services through the port Dr through through we have zero RPODR as well as a DR for replication.
29:31
So we have DR services in the platform. Then you know, we have continued to add more and more services in the port platform. We have port data services, which is our DV services offering where you know we have, you could go select any database point, your communities cluster anywhere with a single click, you can deploy the database fully in production and we manage the life cycle of the
29:52
database. We manage everything for the database with respect to upgrading patching underlying infrastructure, briding upgrading the database we curate it. So we have like a close about 12 databases supported. Today. We're continuing to expand that and we have, we are continuing to add services into it.
30:08
And of course, the most, the hottest topic today these days is A I right. I mean, port has always enabled customers to run A I models uh for a long time because when communities took off and containers took off A IA A models and all of those were kind of the modern workloads and they were all automatically containerized right away. You know, if you do know open A I runs 7000 communities nodes to build their large language
30:33
models, right? And this has been always an arm. So for the last few years, so port works has been always involved in A I development. So we're kind of we also have ability to run a workloads on port works. So the platform has to support all of this, you know, it needs to create a secure environment, it needs to provide provide reliability, business continuity availability and then
30:54
support a whole slew of applications to be run on it. And that is what the platform delivers. So combine that with you see the complete self service platform experience. Uh And that is what the platform engineering teams need today, right? In order for them to be able to go and support
31:13
a large group of developers and then help them build applications and ship them at rapid speed and continuing to update them. They are to marry communities in port together. And with that, what do you get with it? Right. When you get it, you get the whole bunch of benefits, but I'll probably, you know,
31:31
highlight a few, couple of them. Of course, you know, it gives you self service, automation and everything. Uh it, you know, all you, you give the other key aspect of it is that it gives you agility across any infrastructure with port works. You can build your application in your data center, run it,
31:49
you can run another port works cluster in the cloud and you can tier or migrate your application state fully with and sometimes with uh no failure with no failure or no no down time, you can do a live failure after applications to the cloud and you can bring them back too, right? So you can you can go both ways and that is you know, those are the kind of capabilities that port comes built in.
32:10
So you can do you know blue green deployments, you can do carry testing of your data, you know, and you can build new developer workflows. You can do test DEV you can snap a name space clone it in the in the same cluster. You can make it available for your test team to you know, run tests on product on on a copy of the production data,
32:30
not the production data don't do it. But on a copy of the production data and you can enable, you know the the the application teams to kind of realize more and more value out of the platform. So through the migration capabilities, you know, there are some of the other things we have done is as we ran uh port works in production in a lot of customers. In large scale.
32:48
We realized uh container density is an important topic. So we built uh application Q OS within the platform. That means you can bring a lot of applications, make them resident in the same platform and they can run without an neighbor issue. And you can actually control how much I your application can consume in infrastructure, right?
33:10
That means you can have a fleet of flash arrays and you can run platform on top and you can essentially decoratively say how much, how much IOP and throughput you want to give to each app. Now port works enforces that you can dynamically resize it. You can set policies in autopilot to dynamically, you know, adjust that based on the load and you know, and port works lets you manage your large scale
33:35
clutter clusters with very little user intervention. That means you don't have to staff up a team of 100 platform engineers to run 1000 note cluster and you can do so with less than 10 engineers or sometimes you know, one of our large customers that has 5000 nodes has about seven platform engineers in their team and they don't have anyone dedicated to port works,
33:55
right? It's all managed, pretty much self managed through policy and, you know, on our autonomous policy engine and they set policies and let it run and it manages by itself. And you know, so the other thing we're working on delivering, as you'll see is uh we're, we're adding more and more ability to run more A I workloads uh
34:15
on to port works, especially around generative A I and uh you know, all of the A I models. So that is something that uh you'll see us uh delivered in the future uh as, as the road map evolves as well with that there are a few more sessions. So would you like to highlight? Yeah. Yeah.
34:37
Any questions for either I'm sorry or for uh Kendra, please. Any questions, other question, um What are the KPIS or NB OS that you use to measure success team? See a lot of these things come and not and for a layoff. So, what is your plan that or how are you?
35:09
Yes. So my, I manage infrastructure engineers. So I don't actually uh I'm not necessarily involved with application development and um the devops specifically, um I know as a business, we're really looking to um improve our regression testing. Um And that's not a matter of necessarily missing bugs that were there that we should
35:30
have caught. It's um understanding how changes we made may have affected other things that we're not testing, we're simply not testing certain things and that affects all the way down to my side of the business. So I'm not in, I'm on the side that affects us because when those, if we push those bugs to production to our customers and they find them,
35:51
the answer is typically throw infrastructure at it as if infrastructure is endless. Right? Shocking, right? I know. Um And you know, as I'm under more pressure to reduce my operating cost, which to me, um our business is also very committed to not reducing head counts. So when you're talking about reducing your
36:09
operating cost, head counts off the table, so you have to reduce hardware software cost, right? So throwing hardware at, at bugs as a bug fix, that's, that's not a good long term solution. So, um you know, again, probably not a shocker. There's not always the best relationship between DEV ops and SAS or production.
36:28
Um And so bridging that gap is really important to us right now. Um They're an agile shop. My infrastructure engineers are a I call firefighters. It's do the work that comes in every day and make sure the lights, lights are on. Those are two very different um uh approaches. So trying to bridge that um and help s understand what it means to be a engineer and
36:51
help assess engineer, understand what it means to be a de ops engineer. I don't know if that answers your question but a question as well. So probably mean around drop good question. So are are using to lifecycle there. Uh So we're very early in our
37:26
journey. Um So I don't necessarily have answers to all of those questions. Um We definitely went to Port Works. Um We had just become a pure customer. Um We have a um and as we talked about our beginning, our journey to Cobert, we kept looking for um our application is very state.
37:47
So we were looking for how to manage that um on the volume side to keep the volumes uh static and consistent. Um And the name that kept coming up from a lot of our advisors throughout the industry was Port Works and I think Peer had just bought Port Works at the time. Um So it was a very easy process for us to go to our pr team. And hey guys, you know, we're interested in
38:10
this Port works thing. What can you tell us? Um We're pretty much only using it for the volumes at this point. Um But as you can see, it can do a lot more. Um And we're definitely trying to figure out what that looks like for us. I think there's a lot that can be gained by using the other pieces. Um We haven't even figured out what our dr
38:29
looks like in right now. Basically, if something breaks, we just spin up a brand new one. That's so uh just one right now. Yeah. Um because we're just running the single third party utility um there and we're in the process of developing,
38:48
figuring out what HR P looks like containerized both on Prem and in the cloud, by the way. So we're doing both, um, you know, a KS in the cloud. Um and, and is on site with port works, um and trying to figure out what that looks like. Um And so we just have the one third party tool in, in containerized right now.
39:08
So we don't really have a good dr strategy, but obviously, we cannot containerize hr P without AD R strategy. And so I think the next year or so we'll be delving into a lot of the other pieces of port works to figure out, you know, how that fits into our puzzle to solve those problems for us. Yeah, that's great. Thank you.
39:32
If there are no other questions, I just wanna, you know, um pitch the upcoming sessions here. The one at two o'clock is, you know, greater developer experiences and some of the presenters are actually in the room here. And then at, at the three o'clock hour, we have this very fun role play um game where we,
39:52
it's uh we have a business called Port Works barbecue and we're going to walk through a bunch of different use cases and Port Works product actually shown in a live demo for a company called Port Works Barbecue. I would highly encourage you folks to come to the next two sessions there and you'll get deeper into the product without actually having to um you know, do much of the coding and all this stuff here.
40:14
So watch it in kind of kind of real life as well. So that's happening today and tomorrow, some of the sessions are being repeated as well, but there is the first session. Tomorrow is a very critical one for folks who are pure customers who are on the infrastructure side, understand how infrastructure and platform teams can start to operate together with a
40:34
little bit more of an understanding of the interplay between the between the teams. As we all know, with the transition from, you know, when VMS became big deal, you know, um you know, there was a huge amount of translation and transition. People had to go through within organizations to make that the central part of their organizations with containerization. Now platform engineering is becoming very,
40:55
very critical and so that session will walk through some of those things as well tomorrow. So with that, we'll let you off to your break and see you back at two o'clock for the next session. Thank you all. You can con you can confirm that whatever I is presented, actually the product works in the next session.
