Architecting for Product Success: Navigating Cloud Partnerships and Vendor Commitments

[Dominic]

Direct link to the episode

Have you ever wondered how ad-hoc solutions can evolve into critical dependencies, and how a comprehensive integration platform can streamline your operations? Peter Ngai, Principal Architect at SnapLogic, walks us through his pivotal role in product development and engineering, offering deep insights into understanding customer business processes before implementing technical solutions. Discover the importance of a customer-centric approach in tackling technical challenges and the strategies for building a scalable platform in the fast-paced iPaaS marketplace.

The XKCD comic I mentioned is this one: https://xkcd.com/2347/

The previous episode on establishing a Center of Excellence is here: https://podcasts.apple.com/podcast/the-enterprise-alchemists/id1760113394

Peter also shares some war stories from his time managing tech products amidst internal political hurdles. Learn why setting clients up for long-term success trumps quick fixes and how continuous innovation and adaptability are the keys to success in today's dynamic tech landscape. We delve into the importance of maintaining clear boundaries between different components of product architecture and how SnapLogic supports emerging technologies like AI and LLMs through a modular microservices architecture.

If you too had forgotten about the excellent War Games, it stands up well even today.

Finally, we discuss the strategic considerations in cloud partnerships. Tune in as Peter shares his insights on the pros and cons of committing to specific vendors versus maintaining cloud agnosticism. Find Peter on LinkedIn.

Don’t miss our Integreat events in San Francisco and London for a chance to connect with industry experts and stay ahead in the world of integration and system architecture: https://www.snaplogic.com/integreat-tour

This is the last episode of Season One of the Enterprise Alchemists! But don't worry, we'll be back in a few weeks with more insights from the business end of enterprise IT. Sign up in our RSS feed to be notified when new episodes drop. Like and subscribe, and don't forget to tell your friends!

Chapters

• Enterprise Architectural Challenges and Solutions

Building Success Through Customer Sensitivity

Navigating Complex Architectural Innovations

Architectural Flexibility in Cloud Partnerships

AI Conversations and Integrate Event

 

Transcript

Dominic Wellington: 0:00

Hi and welcome back to the Enterprise Alchemists. I'm Dominic Wellington. I'm here with my co-host, Guy Murphy. Hey, Guy, how are you?

Guy Murphy: 0:06

Good, as ever, Dominic, good, good.

Dominic Wellington: 0:09

And we're also joined by our colleague, Peter Ngai, who is our principal architect around here at SnapLogic, and we thought that it would be good to have a conversation enterprise architects to a different sort of architect about what that world looks like.

Dominic Wellington: 0:28

So thanks for joining us, Peter!

Peter Ngai: 0:29

Sure, Great to be here. Thanks for having me. This is awesome.

Dominic Wellington: 0:33

And so maybe that is actually a good starting point. So why don't you first introduce yourself for the benefits of the audience, so they know where you're coming from, and then we can talk for a minute about those different definitions of architecture?

Peter Ngai: 0:48

Sure thing, sure thing. Yeah, my name is Peter Ngai. I'm principal architect here at SnapLogic, so essentially the architect that I kind of play at SnapLogic falls into product development or engineering, as people say, or the development of the core product itself inside SnapLogic. So a lot of what I do day-to-day is provide technical guidance across engineering, oversee that and also promote operational excellence and basically how to promote engineering practices across the board, how we could all move better, faster, quicker, better and more current-day technologies.

Dominic Wellington: 1:31

So you're architecting the actual product, what we enterprise architects will actually be helping customers to deploy.

Peter Ngai: 1:38

That's right. It's more like the internal structure right rather than the external structure, but yeah, both roles are very important, obviously, and basically, that's how we kind of move the ship forward.

Dominic Wellington: 1:53

Fantastic. So when we're doing that, when we're doing the enterprise architecture piece, when we're out there talking to customers about the product that you have helped our engineers, architects, what do you see as the main areas of focus? What should be the topics that we touch on in that conversation, that you're thinking about when you're architects in the product itself?

Peter Ngai: 2:15

Sure thing. Yeah, when working with customers, there's actually a different tactile feel, so to say. Right, I mean, we're both architects and such in our own right. But dealing with customers also requires another batch of skill sets that one may need to possess right, so to say, because in essence you're dealing with both a different corporate structure right, as well as people right, a different corporate structure right as well as people right.

Peter Ngai: 2:48

So it's very and it becomes a very labyrinth of navigation and how you can do that right, and it's not a very complex problem, but it's a different problem, it's not a technical problem, so to say, and how we can best help customers, right. So the first thing to really do is to really understand. Although you're brought in as a technical resource, it's best to understand their processes first, their business processes and what they're trying to accomplish. Right, and essentially SnapLogic, since we have such a powerful foundation. Then we can figure out how the technology overlays in there and how we can actually help right, because we're brought in for a reason, right, essentially, it all starts with you know, they just wanted to connect a database and maybe, you know, create an Excel spreadsheet from that.

Dominic Wellington: 3:39

That's all they wanted to do on a weekly basis. Like all technology, there's some core problem, some job to be done yeah you have to understand that, otherwise it's no good.

Peter Ngai: 3:49

That's right.

Peter Ngai: 3:49

And then i t becomes like one person's job, and it's not their day job, by the way, you know this. They just kind of said, well, hey, I can code this up, I can basically query this database and throw it into a csv file and put it into this folder. I can do this on a weekly basis. It's not a big deal, but, understood, it's not really their day job. They just became the hero that day at 5 pm on a Friday and they were able to crank out this Python script.

Peter Ngai: 4:17

Well, a couple of weeks have passed and now people are asking for more. Well, hey, you know developer, know developer a can you know you know what to be great if we can just add these extra columns and these extra database tables. You know, it'd be wonderful if, just in case it doesn't work, just in case those aren't there, let's take a different route and maybe your script can also do this other thing. So, next thing, we know like it's, you know, the person's updated the script and now they're getting more demands on the script to update it for facilitating different routes and different scenarios. What if the database is not able to connect? What if we add a new table. What if we want a different spreadsheet? What if we want different reports for different people?

Dominic Wellington: 5:02

Right, and now it's a dependency for a bunch of core business services and you're living in the XKCD comic with the one little thing holding up the whole stack.

Peter Ngai: 5:10

That's right, that's right, and so, and all along, let's say that you know this is a couple weeks down the road. It was no one's day job to really handle this process or handle this process and keep in mind, now it's become critical, now it's a critical workload right for the company.

Dominic Wellington: 5:28

In the old days that sort of thing, the problem you'd have would be this it would be running on a repurposed desktop that was hidden under some developer's desk and nobody in IT knew about. That's right.

Peter Ngai: 5:40

And these days it runs on Raspberry Pis, which is fine.

Guy Murphy: 5:43

But, essentially that's correct.

Peter Ngai: 5:47

And so along comes SnapLogic. Now, why do customers acquire such an iPad solution? Well, it's because they've kind of done a due diligence and they've weighed out cost versus effort, versus what they actually get at the end. And companies today, they're very smart Cost versus effort versus what they actually get at the end, right, and companies today, they're very smart, right. Essentially, there's two purposes, right.

Peter Ngai: 6:12

And it's been going on for a couple of years, even before the AI trend kicked up. Right, it was all about cost savings, right. You know, we were winding down a bit. You know, globally, economically, things weren't looking all that great. Things were just kind of pausing. People didn't know exactly where we were headed. But companies had one thing in mind they needed to save funds, they needed to save on expenses.

Peter Ngai: 6:40

Solutions like SnapLogic become very vital because, all in all, after you do the math, it becomes a lot more efficient, it's a lot cheaper, it's just a more elegant solution that one would have an iPass platform that they can integrate on their systems, rather than, you know, hiring a whole suite of developers or, you know, have someone's evening job, keep updating that script because it just wasn't scalable, right. So we kind of have to have that lens or have that sympathy towards the customers, that this is the first reason why we're in the house, right, Essentially, is to help them save money, to become more efficient with their processes and then to be able to scale out accordingly, right. So that is like the other sense, that's the I think a third or a second or a third sensitivity I lost count or for enterprise architecture, right, or enterprise architects once you walk in the door for the customers, this is what they're feeling, right, they understand the technical challenges they have and that's why you're there first and foremost. However, behind the scenes and all the soft skills and thought processes behind it are all about the other points that we have to understand right. And so, with that sensitivity, enterprise architecture, how we can actually bring value right to the customers as we walk in the door, is to have that empathy, to have that lens and be prepared for that, the scenarios and the conversations to have right.

Peter Ngai: 8:18

And myself, you know I have a consulting background, obviously not with SnapLogic, but in my previous life I've been around for 20-somewhat years and I've had my share of war stories, so to say, and every place is unique and, as a consultant, when I was brought into a place. It was always they needed me. Yesterday. It's always the case, right? It's not like oh, you know, we have this great initiative coming up. Unless they were well-planned, right, you know, we'll need like 15 consultants on board and welcome.

Peter Ngai: 8:56

Peter, Welcome to day one. And you know we're going to here's your breakfast and you're going to work from, you know, eight to four or nine to five, and we're going to have a cherry day, right, you know, it was never like that.

Peter Ngai: 9:07

Now it was always like we needed you yesterday. This is what needs to be fixed, this fire and this fire is happening this week and we need to push things forward. Right, but just to have that sensitivity and that lens of working with customers is vital, right, and I think it's the best way to really offer the help that they do need and it benefits them tremendously, right, so it's not like we're kind of in the door, we fix something, integrate a couple of systems and we leave and we kind of shake hands and we're gone. No, that's not exactly it. It's like we need to set them up for success, scaling forward. Right, to make all their processes with the lens of moving forward, better, right, and, and, and that's essentially what, what, the what, what the story is Right, and I there's a couple of war stories I've had in the past.

Peter Ngai: 10:03

It's kind of interesting, but like. One that comes to mind was, essentially, I used to consult at Cisco and I used to be an electronic records management specialist, right, essentially, I would integrate different electronic record systems, records management systems, excuse me with their content repositories, right. So, essentially what. The deal there was that if a discovery order came down from the courts, let's say for an Enron case or something like that. Basically, they needed to provide all the evidence about Enron Not to sound, I just made that up, but just using Enron as an example. They would need to produce all the documents pertaining to Enron from their content repositories and then present that to the courts, like in a couple of weeks. Well, how do you do that in an automated system?

Peter Ngai: 10:55

At the time when I was working there, I think there was a huge, it was like a small city, it was like over 300,000 people, right, and basically the department that brought us in was in the it side of the house for cisco uh, they're. At the time they were divided into it and engineering those two sides of of the company and this was on the it side of things and it was actually this, this one department I'm trying to call from memory, but I believe it was called tax and finance, right. So you know they were obviously well-loved. I'm being sarcastic Fun group of people yeah.

Peter Ngai: 11:35

Well-loved across the whole company, right? So, as you can imagine, one being new to the company, two being a consultant to the company, two being a consultant, and then three having to work with, majorly the entirety of Cisco and their content repositories, along all these groups, and then suddenly knocking on doors and say, hey, listen, I'm the new tax man and I need to put controls on your content repository, just so we can produce all these documents in case of legal matters. Right, I was not well received, right? So, obviously, that just to traverse those waters across the company and across all those departments, across all the political red tape, and such was, was, um, you know, it was pretty, it was pretty challenging, I have to admit, but it was. You know, at the end of the day it was very fulfilling. I would have to say, right, obviously, as a technical challenge, to go there and put um controls under content repositories was simple. It was was a technically simple solution, right, Because it was already built. I just needed access.

Dominic Wellington: 12:49

But as usual the problem is those pesky humans. But this also bridges nicely to the conversation we had on the last episode. We had Chris Ward on and we're talking about centers of excellence and how to establish those on. And we're talking about centers of excellence and how to establish those. No-transcript, that's right.

Guy Murphy: 13:11

Absolutely. so

Guy Murphy: 13:13

So peter Peter, Obviously obviously the main focus of your day job is our internal architecture and how we build a successful, scalable platform . Because, again in the context of the iPaaS marketplace, we've been around quite a long time, but we're also very fast moving and we're also accelerating our change cycle. How do you approach the design and the operations from your role of our platform to cope with our never-ending drive to innovate and bring new capabilities to market?

Peter Ngai: 13:42

Yeah, sure thing. Well, first and foremost, I'm very sensitive to basically how our software is deployed right, essentially so, behind the scenes, without getting too deep. Essentially, if you look at two parts of this, there's one it's considered the control plane which is a central multi-tenant system right, that we all use and that's where customers log into. Right, that's the user interface and such and much of that is there. Much of our data stores are there as well, that are central right. That's the user interface and such and much of that is there. Much of our data stores are there as well that are central right. And then the second component, which is highly sensitive, I think, is where it's called the SnapPlexes.

Peter Ngai: 14:27

Now, there's a couple of different flavors of that, although the software would be the same right. There's essentially what we call CloudP lex, which is deployed into the cloud that we help manage, and then the GroundP lex, where we provide the software to the customers and they can deploy it on premise in their cloud whatever they may want. And so the sensitivity there is just to understand the boundaries between the two systems and how they communicate and basically to honor that distinction right, because I mean, it's one thing to be deployed into like another startup that may not have strict security boundaries and therefore the lines get blurred between control plane and Snaplex and people wonder, hey, why isn't it all in one component? What's the big deal, right? However, when you do get deployed into other, especially Fortune 500 companies, especially in different industries like finance and such and manufacturing even, it's very, very important to understand those boundaries and basically district guidelines, because not one system can always talk to the other one. You know, all day they may, or even in the strictest sense, just to give you, um, a fictitious, uh, maybe one day might be one of our scenarios as well.

Peter Ngai: 15:49

It's like, if we're deployed onto, like a train, so to say, our SnapPlexes are deployed onto trains and obviously we have the control plane that's multi-tenant and central to everybody. But if the SnapPlex is deployed onto a train, let's say on-premise, so to say, but that train is also traversing Europe, right, it's going across Europe, it's on a Eurorail and such, and just keep in mind that there's not constant internet capability, there's not constant internet communications, right? So let's say, the train now even has communication with a control plane only at certain stops through Europe. Now, what happens? Right, it has to be able to be fully running in between all the time the snap plexus right. So there's a lot of nuances there to be sensitive towards and although that's not a new capabilities and trudge forward, essentially not to kind of pin ourselves into a corner, so to say, and find that as an impossible scenario. So we essentially need to kind of keep certain things in mind there.

Guy Murphy: 17:14

Peter, as our architect. I mean, we've got a very powerful platform that is an iPaaS, so cloud-based, and we support multiple integration capabilities. When you approach thinking about the architecture and the roadmaps, what are the key things that you need to consider to not only maintain what we do today in a structured, stable manner, but also be allowed to innovate? So, if you look at the company, in the last two years we've introduced a whole suite of API capabilities that are rich and powerful but are very close to the traditional model of what we've done before, and yet in the last nine months, we've started moving into the AI space.

Guy Murphy: 18:03

Well we're starting to see very different patterns and usage. So, when you think about our landscape, how are you thinking about this constantly changing landscape? Where we spent several years incrementally focusing on traditional integration patterns in the cloud, now we're moving into very different domains, different uses. I know that we're actually thinking about having to change the platform, so how do you approach architecting our platform to keep on supporting multiple use cases and multiple capabilities and products?

Peter Ngai: 18:41

Yeah, sure thing, sure thing, um, so, so, um, and that's a really great question, because the landscape of technology changes ever so quickly and as time moves on, it's even quicker, right, so to say, everything that happened in this last six months previous to that, when I mean, obviously, when we progressed from cassette tapes to cds, that was one thing, but cds, mp3s, that's another thing. And time and technology progresses quicker every day. It's a very exponential equation, right?

Guy Murphy: 19:14

I am going to admit, though, that I've gone retro and I'm building up a very nice vinyl collection. And, there are customers out there that will also be quite retro.

Peter Ngai: 19:24

Oh, that's great, that's great. And, yeah, let's not forget them. I mean because there's great value in being retro. Right, I myself too am in that same boat, but yeah, so critical capabilities moving forward, and we could just take as an example.

Peter Ngai: 19:41

So it's best to really structure our software components internally to SnapLogic platform into microservices. Right, so to say, because the old school where everything was monolithic, has kind of come and gone. Right, because we now understand that the world moves a lot quicker and it's really difficult to, you know, update a couple lines of code and pull, turn around and deploy it quickly and have it accommodate the world. Right, we can't do that, we can't do that right, and so essentially, what we really need to do is component eyes, our software internally, right, so when new capabilities do come up and we need to, let's say, communicate with LLMs, how do we do that? Right? So how do we do it so quickly? It's basically how we structure our software components internally right, to be able to do that, to be able to move at high velocities. Right, where basically one side of the house can stay stable. On the right wing of the house we can actually add, you know, in addition to the house that talks to an LLM, so to say right, and it can then progress the whole platform forward as well. So I think that's a very and the whole story about AI is very compelling right. That actually helps, I feel like the efficiencies for customers as well. Right, because not only, like today, to have a lens towards a customer to say, like you know, they primarily want to save money and they need to scale forward and be more elegant. You know, they're looking at AI to actually help promote those efficiencies as well, right, and so it is our duty to basically help in that sense, right.

Peter Ngai: 21:33

And how can SnapLogic do that? I mean, AI is a very controversial topic, right, understood, depending on how we talk about it. What industry you're in uh, it, it's. It's highly complex, right, and it can be controversial. SnapL ogic being a tech company. We are a technology company where we're enabling you know the use of this, yes, we are right. And so I think, just to kind of speak on this, because it is such a sensitive topic, I mean, the day will come when there is some governance around it where things are. You know, we're not trying to replace jobs, right, we're not trying to, you know, build out. What was that war? War, was it Tron? No, it was some old movie where the computers kind of took out the world and everything. They played some war games or something where they'd simulate it. War games.

Dominic Wellington: 22:30

It was War Games!

Guy Murphy: 22:31

It was called War Games.

Peter Ngai: 22:33

Yeah, so essentially you know we need to govern against that as well, and that's well understood. But I think a lot of that comes from that can be fearful conversation as well, right, just like the turn of the Industrial Revolution, where machines kind of came in and made things more efficient. Where machines kind of came in and made things more efficient, were able to produce a lot more products more quickly. But the biggest fear at the time was basically replacing manual labor with machine labor. So it feels like we're at that cusp again and yeah.

Dominic Wellington: 23:17

I mean, the angle that I like to think about is how there's this distinction in ops circles between the the actual work that you're doing and the toil, the stuff you just have to get through in order to do the work, to be productive. And, as you were saying at the beginning, SnapL ogic's entire proposition has been to remove the toil and let people get on with their actual jobs, with what they're actually trying to do, rather than maintain a tottering pile of Python and shell scripts and what not that they were previously using to pass information around. And so if we can do that with a no- code graphical user interface, that more people can do that with a no-code graphical user interface that more people can use, that's great. And AI is just the next step along that direction.

Guy Murphy: 24:07

All right, absolutely. so

Guy Murphy: 24:12

So Peter. I know we've had a couple of conversations about this and if our listeners will bear of conversations about this and if our lessons will bear on this like yourself, I've been around the industry for a long time and we're working very closely with AWS. I'd love to get your points of view on that. There's been waves over the years of your career, my career of should you bind yourself to a back in the day, a database or a vendor or a technology stack? And now, obviously, with the cloud platforms out there, there's um, the capability has changed drastically. Um, you know, we previously talked about the fact that if you looked at where, when aws and azure launched, um, they were very, yes, and especially from an infrastructure point of view, and this is no disrespect to any of the cloud platforms but back in the day, you really were just effectively buying a virtual, well-managed data center, whereas now, today is very different. They've got hundreds or thousands of capabilities, they've got deep and, in some cases, unique IP and capabilities around the platform and beyond that. And, as I say, we're in the public domain as a partner of AWS.

Guy Murphy: 25:29

But I'd love to get your views on that pros and cons of making a decision as a platform architect, not just at SnapLogic. On that balancing act of the old, I'd like to be portable Vs. I'm going to maximize the capability of a particular vendor and I've sat in workshops with you and I'll be honest, I was very impressed with AWS's depth of knowledge beyond just the technical stack into actually being a good strategic advisor. But I also had that slight kind of not negative reaction, but that oh. But we are almost committing. If we go down this path to if we maximize on one approach, we are also um tying ourselves to their way of thinking, which may or may not be right with their knowledge of their platform. But this is a question I know that many architects have when they're looking at this sort of level of how much do you really tie into the value proposition of the big players nowadays?

Peter Ngai: 26:31

Yeah, that's a great topic and just to say, first and foremost, AWS, we do work with them as a partner and they've been wonderful partners, right. Just to say that first and foremost and that comes from different facets too, but I really think it stems from their value system, right. So, like, if you kind of like read Amazon's's like values and I don't work for amazon, but I believe there were like something like 16 of them or something like that but um, but they clearly do have the customers um interest top priority, right. So that is first and foremost, um, and and just to, to come from that place of uh, being that genuine right, with that high integrity, it in it by itself is wonderful, right, and and and it's just great to work with them as partners and um.

Peter Ngai: 27:34

Coming to the question of architecture, yes, of course, as an architect, we just feel it in our bloodstreams to be open and not pin ourselves into a corner, and for any technologies that we do choose, we need to stay flexible, we need to be able to accommodate, we need to be essentially cloud agnostic. To be more specific about the AWS conversation, we need to be cloud agnostic. We need to be able to also integrate with GCP, azure. Whatever you may have private data centers, we need to be that open and when a choice does come down to whether you choose a certain technology that may be open source or something that may be just AWS only, even though it provides you the same feature set and we've been there, we've been there Just to give you a drawing from experience working with AWS our NoSQL data stores right when we had to. Essentially we wanted to go managed right. So essentially we and hopefully that's okay to talk about that in here but we wanted to go managed right and our choices were essentially do we go MongoDB Atlas, MongoDB Cloud, so to say, right, or do we go AWS specific and do DocumentDB right? You know AWS being good partners and providing information, of course they brought DocumentDB to the table. Of course, why not? It's their product, they're proud of it. Right, it's fully capable, has a lot of features that one can really benefit customers, right? So of course they're gonna bring that to the table. But with that they brought it. They brought Document DB to the table with that at heart, saying this is what all the features you can use it for and this is how you can expand out and this is how we can tell you if it is right for you right, and so essentially we went down that route.

Peter Ngai: 29:42

We went down that road both at the same time with AWS and MongoDB Atlas came out clearly ahead for our needs. We needed that and AWS was totally fine with that. They understood and that's how we value their relationship, because they are that they do hold such high integrity, knowing that it is yes for SnapLogic, it was the best interest for them to stick with Mongo and go Mongo Atlas for managed. So that's wonderful, right, it was like. You know, I have worked with other vendors in the past for different companies. That was not always the case. It was always like, not always quite so pragmatic. Yeah, no, I've run across the salesperson who's I don't know was pretty sour that they lost the deal and although I went to them and said, hey, listen, you didn't get the deal, but you know I'll keep you in mind for other stuff and I'll refer you as well if I feel that it's right and that was not well received for that salesperson.

Peter Ngai: 30:54

I don't know, maybe they just maybe their rent or their mortgage was on the line for that needed to do something with you, yeah, I don't know.

Dominic Wellington: 31:02

Something that is always very interesting on the from a product side is you know, sometimes it happens you lose a deal and you're like okay, but why? What was it that we were missing? What was it that that prospect was looking for that we were not able to provide? So that we can figure out. Okay, is this something we might ever do? We want to put it in the roadmap so that next time we run into this situation we are able to do it. Or was this just not a good fit? It's something that we don't do, we're not planning to do. If we see a situation like this, we should walk away sooner and not waste everyone's time.

Peter Ngai: 31:37

Exactly, and I understand, should walk away sooner and not waste everyone's time. Exactly exactly, and and it's um and I, I understand, I get it right. I mean, this is essentially a a business that we're running and sometimes it crosses the line and becomes personal. For some people, maybe it does come down to some personal financial matters that they needed the deal, they needed the commission. I understand, I empathize um, and at the same time, I try to. I try to detach from that because it is, it isn't anything personal in this case or it's not coming from a personal place, right, so but yeah, yeah, sometimes we we go there.

Peter Ngai: 32:15

We don't want to go there.

Dominic Wellington: 32:19

As we said at the beginning, the whole human factors thing, it always comes up. Just in a few more minutes, there was one more topic that we brainstormed that I wanted to address quickly. So we're seeing a little bit of a trend, and we were just talking about cloud. We're starting to see a little bit of a trend and this, again, is one of those things that comes back every year a couple of years. But talking about cloud repatriation and the data center isn't dead after all, and we on this call definitely all knew that I work with customers who still have mainframes, for goodness sake, and when were they supposed to die again? I forget.

Peter Ngai: 32:59

But how much of care Exactly?

Dominic Wellington: 33:02

How much of that are you seeing and thinking of, and what would you tell people to consider when they're looking at that topic?

Peter Ngai: 33:13

Sure, sure, and essentially, when we design or architect our products, we don't assume, we assume the software right, essentially, so it can be deployable to both you know, someone's desktop, so to say, or laptop by these days and also can be deployed into the cloud as well. Right, so we stay agnostic to that fact, right, and there is again, there's always that pendulum that swings between, you know, hardware data centers and cloud, private data centers. Is that a different clock? No, that is just another person's data center. So that pendulum swings back and forth, right, and depending, it really depends on the customer and what they really need, right, a lot of times we may even, you know, right now, the most efficient way, the most expeditious way that a customer can get on board, if they really wanted a cloud, a groundplex, excuse me, we provide them the SnapPlex software is that they can deploy it into the cloud. They can spin up an EC2 instance in AWS, right, throw it on there, of course, right, they can also start up the Docker image on their laptop that they're writing the email on and they can also run a SnapPlex on there.

Peter Ngai: 34:39

Sure, that's too. But the point is, is that that we need to stay agnostic, right, and then essentially accommodate both and move forward. And then essentially, how can we run best in all scenarios and how do you accommodate that Now? Does it mean like everything is restricted down to the nth degree? No, it doesn't mean that we can. Actually, we should be able to run in different types of configurations, right, depending on the deployment that is needed.

Dominic Wellington: 35:11

Yeah, meet the customer where they're at, depending on what the famous business need is.

Peter Ngai: 35:15

Yeah exactly.

Dominic Wellington: 35:19

Great, I think that's a fantastic point to wrap up on. So thank you, Peter, You've been a fantastic guest. So that's the end of this episode. It's also the end of the first series of the Enterprise Alchemist podcast, so thank you for those who've listened to us so far. If someone sent you just this episode, by all means please go back and listen to the others. They are not topic of the week type of conversations. They still stand up. We mentioned Chris Ward's conversation about Center of Excellence, but we also had a bunch of really good conversations about various aspects of AI something in there for everyone in the enterprise architecture space, I would say. And if these topics are interesting, you should also join us at SnapLogic's Integreat event. We're running two of those. One is in San Francisco on the 23rd of October.

Peter Ngai: 36:11

I will be there.

Dominic Wellington: 36:13

Fantastic, and Guy and I will be at the London event, which is on the 7th of November. We will put the link to register into the show notes, but in the meantime like and subscribe, because that way you'll also be notified when season two drops with even more exciting topics and guests. We look forward to talking to you then.

Peter Ngai: 36:36

Absolutely. Thanks for having me. It's such an honor too, and if you happen to be at Integreat, if you listen to this, please stop and say hello to me, and it'd be wonderful to chat. So thank you and thanks for having me on. This is wonderful.

Dominic Wellington: 36:50

Absolute pleasure. You've been a wonderful guest, thank you.