Simon Margolis (Director of Cloud Adoption at SADA) flexes his extensive experience with the Google Cloud Platform, and offers key information on the potentially incredible future of data, AI and map technologies.

Episode Transcript:

INTRO: [00:00] Welcome to the Tech Deep Dive podcast, where we let our inner nerd come out and have fun getting into the weeds on all things tech. At Clarksys, we believe tech should make your life better, searching Google is a waste of time, and the right vendor is often one you haven’t heard of before. 

Max: [00.18] Hi I’m Max Clark and I’m talking with Simon Margolis, who’s the Director of Cloud Adoption for SADA Systems. Simon, thanks for joining.

Simon: [00.25] Yeah, absolutely – happy to be here.

Max: [00.27] Simon, we should start talking a little bit about SADA Systems sand what SATA Systems is, so can you give me a quick thirty-second-ish overview of what SADA does for customers?

Simon: [00.39] Yeah so, high level: we are the world’s largest reseller of the Google Cloud Platform, that includes both our resale business as well as our systems integration business, where we provide professional services for our customers to aid their adoption and maximize their use of Google Cloud Platform as a whole.

Max: [01.00] And you personally, before you got into the cloud world… I noticed a note about a very interesting company out here in Pasadena, and it looks like – I’d love to talk to you about that a little bit.

Simon: [01.09] Yeah! Yeah, I mean I still have to kind of pinch myself sometimes to believe that I really got to do this, but I did – before I was at SADA – get to work at JPL, NASA JPL out in Pasadena, which was an amazing experience, I highly recommend it to anybody who has a chance to do that. But yeah, we did a lot of systems work there, that’s kind of the thing that opened my eyes to cloud computing, because it was very early days of AWS back then when I was at the lab, and of course at NASA they encourage experimentation, trial and error, playing around with new tech… But that’s an experience I will always cherish, a very cool place to work.

Max: [01.50] I mean, just thinking about the nerd factor of it – I’m already there! I mean, so NASA and JPL were very progressive and – use of Linux, creation of physical hardware, massive storage servers, and then you know, different nacient cloud technologies and compute, farms, commodity infrastructure; you’ve quite literally seen a big transition in here in not a long period of time, from physical on-premise to cloud, and that’s an interesting… It’s an interesting seat and view of the world.

Simon: [02.24] Yeah, no doubt.

Max: [02.25] Google Cloud is dominantely known for their G-suite, and SADA does have a practice around G-suite – that’s not what we’re here to talk about today – but Google created the Google Cloud Platform, and they have a somewhat of a different slant on how they’re approaching things, from Amazon AWS and from Microsoft Azure. You know, from the SADA view, from your world, how do you view the differences between this platforms, and what GCP is bringing to an enterprise that is different or unique from the others?

Simon: [02.27] Yeah, it’s a really good question, and it’s at the core of what I’m trying to solve for at SADA. I think that both – at SADA – that we have a reputation for being a big G-suite partner, and I’m trying to change that, to make us known also for being a big GCP partner. The other thing is that Google is sort of seen as being this – I guess for lack of a better term – LADDERED(?) in the cloud space, and a big thing I’m trying to change is that perception. I think a lot of the reason they’ve earned that perception is because of their approach to the market, and it is different than what AWS and what Azure did, and I think there are benefits and disadvantages both to their approach, but you know, where I’m coming at this from – like we just sort of talked about, right – it’s the fact that… At the end of the day, I consider myself a nerd; I don’t think that’s a negative term, right? It’s something that I wear proudly! I think there’s a lot of folks like me who are, you know, into the technology and want to use the best technology, and I think – I’ll poorly paraphrase Eric Schmidt when I saw him at one of the early Google Next Cloud conferences, but he basically said: “Look, we built App Engine, right? Platform as a Service, you used to upload your code and away you go, like Google do with the hard stuff. What more do you people want?” And of course, at the time I was like, “Exactly, right? Of course! This is all one would ever need, right?” But you know, I think we nerds maybe didn’t have an appreciation for where enterprises were in their adoption of these emerging technologies, and so what made sense to a young engineer, early in his career, maybe was not the thing that your Fortune 500 organization was going to adopt. And so, I think that’s at the core of the difference between how Google has gone to market with their cloud platform versus how some of their competitors have.

Max: [04.50] Google was definitely early with App Engine; I mean nowadays we talk about ‘serverless’, right? When App Engine came out, serverless wasn’t a term, support for it wasn’t really there, Google required a very specific data interaction, data layer, that was abnormal for a lot of people used to working with SQL engines, and they were definitely early. But I mean, the other side of it is Google and Google Cloud has given the world a lot of things that everybody knows… Containerization and Kubernetes, a machine learning pipeline. So, are you seeing that as being a big differentiator when people are looking at which cloud platform, or cloud switches, or multi-cloud… You know, is that driving this conversation, or what brings people in to Google Cloud today, and how does that start?

Simon: [05.43] I think you’ve got it on the head. I think there… So you know, I’ve obviously – I’m a little bit biased here in the sense that I’ve tied my career to the Google Cloud Platform, so that’s probably worth mentioning, but I think that is exactly it: they were early in their technology, and we’re starting to see that paying off now in the sense that… You know, I remember the very first Google Cloud conference I wnet to, where they were like, “Hey, there’s this thing called Kubernetes,” and we think it’s going to be this big deal, and everyone is like, “What?! We’re VMware, what are you talking about containers for? It doesn’t make any sense!” And so they were way too early. Now, absolutely – this is the reason that we’re having conversations with clients about adopting the platform, and containerization is one such unique technology that Google has that is attractive to clients today, but you have to also remember that Google was Google.com before it was anything else, and as a result the data technology that Google has, they needed to build in order to provide a Google.com experience – is something that they’ve continued to keep their foot on the gas pedal with. And so, as a result, I think they had a lot of technology in terms of data analytics and data warehousing that was far too advanced for most customers five years ago, and because of the cost of storage going down, and because the accessibility to large volumes of storage… Today, that’s really not the case – it’s not too early any more, and so yeah – we are definitely seeing those unique Google features being the things that are leading customers to open their minds to using Google Cloud Platform.

Max: [07.26] So I mean, beyond Kubernetes and containers, tell me more about this. I mean, what are the dominant Google features in GCP that are driving customers to Google and that, you know, you’re having the most conversations around?

Simon: [07.42] A personal favoruite of mine that I think is a bit underrated is Google Big Query. It’s another one of those early products: Big Query has been around for a while, probably one of hte older products in Google Cloud Platform, but I think it’s like App Engine in the sense that it was way ahead of its time, in terms of its power and ease of use, but that’s a technology that we find many customers taking advantage of, across the spectrum. This isn’t just your Bay Area startup that’s focused on data science that’s using this stuff, rather we do have enterprises that are moving to… Using Big query, where they otherwise maybe wouldn’t be adopting any emerging technology, just because it’s so powerful, it’s so cost effective; it’s a great tool. There’s other – maybe slightly nerdier – components around the data engineering toolset that the Google Cloud Platform has, that are also tings that are really leading our customers to adopt. The thing about that is, and that’s why I love doing things ike this where where we can talk about the various products Google has, is because I think a lot of customers out there – especially the enterprises – aren’t aware sometimes of what may be out there in the Google ecosystem that may serve their purpose. An example of this is, we do have a lot of conversations with the customers around Big Query, however we find that some customers, their needs are not met by Big Query. It leaves them asking for more, or a differentiator, or something like that, and in the case that performance on Big Query isn’t what it needs to be, customers who want – you know – single digit millisecond response times, they may not know that Google has a product called Big Table, which does that! It does deliver those single-millisecond response times, from reading and writes, you know? I think there’s going to be another App Engine, and there’s going to be another Big Query as customers learn more and more about what the capabilities are on the platform.

Max: [09.32] Let’s talk about Big Query specifically for a little bit, and this is something where a company that already has a MapReduce or Hadoop-based workflow ends up looking at Big Query at some point; the pipeline for that job, and the analytics for that applies quite well, right? But now Big Query is very different in how it prices, and we have these things called slots. So how do slots apply, and why are people driving towards Big Query so quickly and consuming this so fast?

Simon: [10.05] So I’ll start with what you mentioned about slots, because it’s an interesting feature to what Big Query offers, in the sense that you can sort of ‘purchase’ reserved units of comptunig, so to speak, where that’s the slot of computing – that’s the thing that makes it a little bit easier to get predictabililty, both in terms of performance and cost on Big Query. And of course, as more and more enterprise customers are adopting these types of differentiated technologies, those types of things become important, in terms of predictability. But one of the coolest things in my opinion, about Big Query, it actually kind of goes against that, in the sense that… Well, just like with App Engine, the thing that is most attractive about App Engine is that you sort of upload your code as the developer, and you kind of walk away. You wipe your hands, you walk away, because Google handles the scale, it handles managing the traffic, managing the traffic, making sure that the customer experience is what it needs to be. I feel the same way about Big Query; I feel that – I would never consider myself a data scientist – but I’m able to do a lot with data that I otherwise wouldn’t be able to do, because Big Query just makes it automagic, it just takes care of these things for me. So, I really enjoy the fact that I can write a couple of queries against my dataset and say, “Okay, that gives me the information I want, and now I can go from one or two queries to tens of thousands of queries a day,” and I know that I don’t need to do anything to make that happen, right? It’s going to automagically scale up to provide the performance that is required to answer those queries, and then the best part is, it then scales back down when those queries aren’t happening. So as a result – unlike with traditional systems – but also unlike some cloud systems that are used for similar workloads, I don’t need to think about nor do I need to pay for resources that I’m not using. And that’s not one of those things where, “Hey, we’re going to scale down to the best of our abilities -” no, Google doesn’t price Big Query like that. Google prices Big Query based on how many bytes you process, and so if it takes a little longer or a little quicker to process your query, that’s not your problem, and you’re not being charged for hte length that it’s taking – or the time that it’s taking I should say – to process your query, you’re just paying for that query. And it’s really beneficial in the sense that I know my costs are going to have a little more control over those costs, but also when you think about that pricing model… Google is incentibized to get my query executed as quickly as possible, because I’m paying what I’m paying, whether it gets done quickly or not, and the quicker they get my query done, the more compute capacity it available for other customers. So truly this win-win, and I feel like those types of technologies: App Engine, Big Query, those are the big differentiators, those are the things that… Maybe when you’re talking to the CTO you don’t hear this, but when you’re talking to a developer they say, “Wow, using Big Query really changed how I work, or changed how productive I am,” and that’s exciting for us.

Max: [12.56] I’ve seen pretty radical pricing differences also for migrations into Big Query, you know, applications are sitting either on-premise or in another cloud in some sort of MapReduce pipeline that went to Big Query, and you know, lopped sixty percent off their monthly bill for that. That’s a pretty compelling argument outside of the tech for why you should look at Big Query, and having you know, fixed capacity in terms of slots, and a lot of cases it was, “Well we know how much, how many queries we want to run per day or per interval, and we’re okay if some take a little longer to run or have to be queued out or whatever the ramification was, because it’s efficient for us to know this is exactly what our bill’s going to be each month, based on this component of our data science.” 

Simon: [13.43] Yeah, absolutely, and I think that’s still true, and that’s the cool thing about the way Big Query pricing has been built out, with both slots and the ability to not use slots, is that you can kind of dial in what makes sense for your business, whether that cost control takes precedence, or whether the performance takes precedence, or maybe a mixture of both, right? That’s what’s being unlocked when those workloads get shifted over to Big Query. The thing that I think that gets missed a lot also – which is super important in terms of time and cost savings as well – is that this is much less work for our customers to use. You know, you’re not managing Hadoop, you’re not managing – not only are you not managing the software, you’re not managing the physical layer beneath it. And so as a result, your developers can get that much more done in a day, which has a reeal serious impact on productivity. 

Max: [14.36] Another cool thing about Big Query is it doesn’t require you to use a proprietary Big Query interface in order to manipulate it. I mean, it extends and tehre are other systems that you can interact and connect to Big Query. Talk about that a little bit.

Simon: [14.50] Yeah absolutely, that’s one of hte best parts – we were all really excited when ANC SQL was officially supported on the platform, just because what we talk about – well what we talk about in general when it comes to cloud computing to many customers sounds too good to be true. Big Query is a good example of that though, in the sense that we say, “Hey look, all this stuff’s going to be handled automatically, it’s going to be cheaper, dramatically, than what you’re doing today, it’s also going to be faster, and also better,” and the question is always, “What’s the catch? Do I have to use some kind of special tool or something?” That’s the cool part – the answer’s no! You know, we have customers whose applications were using SQL to interface with their data warehouse, and they still use SQL to interface with Big Query. Look, we’re in the business of doing migrations, so I’m not going to stand here and say it’s – there are very difficult migrations, but look, yes, some are more difficult than others. When it comes to adopting something like Big Query though, it’s extremely simple, because again – there’s just that backwards compatibility, so to speak, with whatever the customer may have been using before and so – unlike with some other emerging technologies – this doesn’t require a dramatic refactoring of thye application, or of the infrastructure itself, we can simply plug in Big Query where maybe another tool had been used previously, which is pretty neat. 

Max: [16.03] And things like Tableau, having a Tableau connector to Big Query gives a massive amount of flexibility and customisation for somebody that’s using Tableau and understands Tableau and has reports written in Tableau just switching where that data and where that processing actually exists from another place over to Google and leveraging Google’s scale and efficiency and underlying technology.

Simon: [16.28] Yeah, it’s huge. I think the fact that you can use Tableau is huge because so many cuistomers are using it today. The fact that you can plug it into anything that speaks SQL is also huge because that means that you’re – all of your third party tooling can still work with Big Query, as it might have worked with something like VC application, and I think Tableau’s a great example, but I think this is what… I think this is a big reason that we saw Google acquire Looker, is to make that integration of visualisation on top of these very powerful data platforms, something that’s more accessible and easier to use. Big Query is a good example of things that are compatible with lots of other services, not all Google tools are so compatible, and as a result, having something built into the platform like Looker, you know, is a really big deal and it really helps bring the power of all this data to the masses, so to speak.

Max: [17.22] You touched on Big Table a little bit earlier. The differences in selection between Big Query and Big Table… Why would somebody end up in Big Query and why would somebody end up in Big Table, or would they use both at the same time? How do you make that choice?

Simon: [17.33] So I think both at the same time is totally something that happens, and that’s – I mean, not to digress too far – but that’s one of the things I like the most about the Google platform, is that I think of it like a set of Legos, and you can take lots of different components and build with each other, and you don’t need to sort of make decisions about, “Do I use this or that?” You can use both. Big Table, for example, you can – I’ll speak in a moment about all the things that it’s great at – but if you want to, you can take your Big Table dataset, and plug it in as an external table to Big Query. And so you know, for example, Big Query is not — I’m sorry, Big Table is not a SQL compliant database, it’s no-SQL. And so we might have customers that say, “Hey, I like all that stuff that Big Table can do, but I want to run reports, and I want to use the power of Big Query, and I want to use my SQL Syntax to run those reports, so I kind of have to pick one, don’t I?” And the answer is no! You can use Big Table for all those cool things that it does, which is all those really fast read and writes, the no-SQL, the schema-less… Makes it very, very fast, but then I can also plug it into Big Query and use all the things I like about Big Query on top of the same dataset. You know, I think that’s the best of both worlds that a lot of our customers are after. But I do think of everything that Google has as sort of a spectrum. You sort of have… You can kind of pick how you want to administer your environment. On the one end, you have tools like Big Query, or like App Engine, that are extremely automate. You know, they really define what a managed service is – you don’t do much of anything other than provide your data, and then ask questions of your data… That’s about it. And then Big Table is sort of step on that spectrum towards less management, in the sense that with Big Table, I’m still responsible for managing the individual instances that provide my Big Table environment, which gives me more control; which often times for our customers is exactly why they’re selecting that technology: they want to control certain components of responsiveness and speed on their data, and so you know, that’s a great example of where you kind of get to pick the right tool for the specific requirements that each individual workload has. In this case, you really get to do it without compromise.

Max: [19.52] Google also has a tool called Tensorflow. What is Tensorflow and how does that apply into an enterprise?

Simon: [20.00] Tensorflow is really cool! Like I mentioned before, and I’ll stand by it, I’m the last thing from a data scientist, and I won’t pretend to be one, but Tensorflow is effectively a platform. It’s a framework, really, that Google has put a lot of energy into, just like a lot of the other open-source projects they’re big contributors towards… And it’s a framework for machine learning, and so – you know, that’s sort of the logical next-step: we talk about Big Query, we talk about Big Table, there’s a lot of ability to capture and analyse data, but then what do you do with that data? And so, the Tensorflow framework really helps with helping organizations build models where they can harness – not just the data that you’re able to store and compute on within GCP – but also the massive computing power that GCP provides, and use those things in concert to build and train custom-built models to answer unique questions for unique workloads. It’s worth mentioning I think, to take that abstract to make it a little bit more concrete, everyone’s kind of familiar with Google’s vision tools, right? If you go to Google.com and you search for an image, they’re pretty good at serving you and image, and for those of you that have played around with it, Google.com also has the ability for me to upload an image and search for a similar image. There’s a really powerful Tensorflow-based model that Google maintains, that allows for that type of technology – you might see that if you used the consumer Google photos app on your phone also, right? You see the same type of thing. And so what Tensorflow does, is it basically allows you to build a model like that for your own use case.

Max: [21.43] The thing that’s always fascinated me with machine learning – we talk about model development – is how much data and computation goes in on the front end of hte pipeline, and how efficient it is on the backside of the pipeline. There was a story a couple of years ago, a fast food restaurant who was using machine learning and vision learning to detect how much chicken they had available in their hot baskets, and make decisions on how much chicken should be cooked based on time of day and predicted load and traffic, and all these sorts of things. And that was a machine learning based pipeline, they were training the model offline and using – a little like, Raspberry Pis or something in the actual franchise location itself to be making these decisions. So it’s an unusual thought process, but it’s a pretty specific example of how do you make a business more efficient from a resource-allocation and capital, time and efficiency… All these things, in a tool that you know, five years ago wasn’t available to the mass market. 

Simon: [22.44] Yeah absolutely, and again – just like with all the different tools we’re talking about that Google has, there’s a spectrum there. And so, I’d say on the least-managed, therefore most complex side of things, you have the Tensorflow framework, where yes – from the ground up, you can build a model, custom to your needs. So, yo uwant to count how much chicken is available, you probably need to build your own model that knows how to identify and count chicken and all those cool things, but on the other extreme end of that spectrum are those pre-built models like we’ve just mentioned, there’s the vision API… The cool thing about that is that you’re able to take advantage of all the things that Google’s constantly adding as a result of their own work on – well frankly, across all of Google. An example is: we worked with a customer not all that long ago that wanted to build a custom model to identify how many people were in their business at any given time, count the individuals, but also get a feel for customer satisfaction. So they were scratching their head, “How do we even build a custom model that will identify those things?” Well, it turns out that the Google Vision API has a pre-built capacity for identifying faces, and so they can easily count the number of people that are there, and it also has a built in capacity for sentiment analysis. So it can look at a face and tell you: is this an angry person, a sad person, a happy person? And so in the case of this client that we were working with, they found that it was much better to simply take advantage of what’s available out of the box – and it truly is available out of the box, right? They simply invoke an API and provide, you know, the imagery to the API and they get all this data back as opposed to building something ground-up, and I think those little things that – you know, in this case – saved thousands of development hours, those things maybe aren’t things that everybody knows about, and they don’t all know that it exists. That’s what makes our job so fun, is that we get to open people’s eyes to a lot of this technology.

Max: [24.45] You’re the Director Cloud Adoption for SADA, who’s a large Google partner, right? I don’t think a lot of people understand the relation between Google and its partners, and where SADA fits into all these things. So, why would a company talk to SADA versus talking to Google? If I’m making this migration and I want to use Tensorflow, or I want to figure out how to use Big Query, how does SADA fit into all that and how do you help?

Simon: [25.08] That’s a great question. That’s been a key question since I started our GCP business over seven years ago now, which is… “I’ve heard of Google but who is SADA Systems,” right? Most folks don’t even know how to pronounce our name, and obviously that’s changed a lot over the years… Winning global awards from Google has really helped our brand recognition but I think we play a fairly vital role in the overall Google go to market, as well as the adoption of the platform… In a sense that it is our job to help customers make use of all this really powerful technology, and you know, the way we see it at least is that it’s Google’s job to make the best of breed technology; to make it, to support it, and to ensure its reliability and maintenance and operation. In concert with that, we feel that it’s our job to work with individual consumers – or potential consumers – of this technology, to help them fit this tech into the problem spaces they need it to be in. So I think that takes a lot of different forms for us – we’re able… We provide support to our customers, to help their applications that run on GCP keep running. We provide consulting in the sense that – kind of like what I just mentioned with that use case, the Vision API, right? We help customers understand that, “Yes, Google makes lots of really cool tools,” we help them figure out which one is applicable to their use case with the least amount of pain and you know, the best end results. And then we’re also responsible for the ongoing success of that customer, right? We’re there to make sure that, you know, it’s fine if you migrate all your stuff to Google – that’s great and we love to help customers do that, but we want to make sure that our customers continue to have a great experience with their use of the cloud, and that goes beyond support, right? I think, again, Google does support their products and they’re responsible for the actual product, we want to make sure that they continue to be used in the best way possible. So, it’s a really good symbiotic relationship I think.

Max: [27.15] And it’s safe to say that Google prefers SADA to be engaged with a customer deploying onto GCP at this point? 

Simon: [27.23] Yeah, I mean… I think that, for the most part, the customers who partner in their adoption of the platform are the most successful both in the immediate – meaning they have the best migration, they hav ethe best success at actually launching their project on the platform, but then they’re also the most satisfied long-tyerm customers. And so for that reason, you know, Google loves to include partners like us, because we’re able to sort of share that responsibility for ensuring the outcome that the customer needs, long-term.

MID-ROLL: [27.54] Hi I’m Max Clark and you’re listening to the Tech Deep Dive podcast. At Clarksys we believe tech should make your life better, searching Google is a wate of time, and the right vendor is often one you haven’t heard of before. With thousands of negotiated contracts, Clarksys has helped hundreds of businesses source and implement the right tech at the right price. If you’re looking for a new vendor and wwant to have peace of mind knowing you’ve made the right decision, visit us at Clarksys.com to schedule an intro call. 

Max: [28.18] Walk me through the cloud journey for a customer. So I make the decision, either I’m evaluating a public cloud vendor at this point, or I’m leaning towards Google… What would that engagement process with a SADA Systems be, and how do you help that enterprise make that decision, “This is where you should go, and these are the applications you should run, adn this is how I make your business do X, Y and Z using these tools that are available”?

Simon: [28.44] It takes a lot of different forms, I will say universally across our customers, when you first engage with us we’re not the type of business that’s going to nickel and dime our customers, and so as a result, we have these types of ideation sessions, we go through their workloads, we go through their requirements, and we understand their environment. We do all that just upfront, right? We’re not – that’s not a service we sell, we do that to better understand our customers’ positioning and what they’re trying to accomplish. But then it varies a little bit, because we do have customers that are already on a cloud platform, maybe it’s Google, maybe it’s something else, and they’re trying to learn what they can do additionally… And for those customers that are already on GCP, often times we’re having conversations with them about what more they can be doing to take advantage of what is already on the platform. We have customers that are not developing anything at the moment. Maybe they’re a startup, and they’re looking for a new platform to start their journey on. And so as a result, we have different sort of approaches for each, but I think most common is the customer that’s migrating to the platform. In those cases, we have a number of tools at our disposal to assess their existing environment, both in terms of the operational efficiency, as well as the actual cost of running, from an opex perspective… And in some cases when we have customers that own their hardware, own their datacenter, we understand what the costs are – we also understand what the headaches are, sitting with that work. Then we’re able to work with them to figure out what makes sense in terms of a move forward, right? Maybe we go through that assessment and we say, “Look, cloud maybe isn’t for you.” I must say  that doesn’t happen very often, but it does happen! But more realistically, we’re able to then say, “Okay, these are maybe some target workloads that we think are good first movers to the cloud.” Or if you’re building net new, here are some things that we think are going to be best to prove the value of the cloud in terms of initial deployments. Then in some cases we’ll work with our customer to actually support that work with them by providing professional services. In other cases we simply will be there as advisors, to sort of hold our customer’s hands through that process. So yeah, it takes a lot of different forms, depending on what the customer needs and where they are today.

Max: [30.55] So another variation of that question might be: I’m already on another cloud platform, and I started using Big Query and Big Table and I started a process to go serverless or I started down a containerization project, and this GCP thing has been over here in the corner with some of our cloud spend, but now maybe it’s a… Should we look at this deeper, because we’re getting into more of the – maybe it’s Google tools, right? So if we’re going containers, and Google literally wrote Kubernetes, which is the dominant orchestration platform for containers, what is that conversation like, and process for people evaluating a switch from something else into a GCP, in a big way?

Simon: [31.39] In that case it’s a little less automated. We don’t have a tool that can say, “Hey, you should containerize your application,” but we will work with that client, and we’ll understand both from a business perspective what it is they want to accomplish, as well as from a technology perspective. An example of this would be… We hear all the time, customers who – exactly what you just said – they want to come to Google because Google is sort of hte grandaddy of containerization technology, they’re hearing containers all over hte place, they’re hearing Kubernetes mentioned all over the place – they want in, right? The first thing we do with those customers is, we don’t dive into their application code, or understanding any of that stuff, we want to understand why – from a business perspective – they want to adopt that technology. What do they want to gain by doing that? I think that’s something that we take very seriously as sort of a core philosophy of how we approach customer problems. You know, we could – frankly, probably make a lot of money by just saying, “Oh, you want to go containers, great! Let’s spin you up a statement of work and start moving you to containers, right?!” But it’s not the best option for every customer, and even more to the point, sometimes what they want to accomplish can be accomplished in other ways that maybe are cheaper or easier, and so that is always where we want to start. We want to know what it is the customer wants to accomplish. The other benefit of approaching it that way is that – like we’ve been talking about – there are a lot of cool technologies that Google has, and when a customer comes in and says, “I really want to become more efficient, and I want to get my developers to spend less time deploying and testing,” maybe the answer isn’t containierzing their application, maybe the answer is retooling their DevOps pipeline. And so, that’s always where we’re going to start, we want to understand what’s at the core of what they’re trying to accomplish. 

Max: [33.27] Also kind of interestingly unique to Google… Google Maps is an incredibly popular application, on people’s cellphones, or they’re interacting with it in a desktop, and this is something that’s available to enterprises through GCP, and something that SADA does a lot of work with customers. What are people using Google Maps for from an enterprise standpoint? How are you helping people with this? What’s kind of common that you’re seeing today?

Simon: [33.51] So there’s a ton of functionality that comes out of the Google Maps ecosystem at Google, and we’ve had – before we had a large GCP business, I think actually before we had any GCP business – we had a massive, I believe at one point we were the world’s largest reseller of Google Maps… And so, there’s been an awesome collboraiton between the application side and the infrastructure side that happens on GCP, and the mapping side of things, and the solutions that we’re able to build for customers range really all over the place. We have customers that are using Google Maps for mapping front-end for their application, right? Think of food delivery, or these types of things that need to have a customer-facing map involved in their application. They’re able to use Google Maps for that. We also have customers that are leveraging things – I’m sure everyone’s familiar with Google Streetview, right? We have many customers take advantage of that, because – and not to go too far of course here, but just to sort of get everyone’s wheels spinning a little bit – think of what we just talked about, that Vision API, and its ability to identify lots of really interesting things in an image, and then pair that with Google Streetview. Then you could sort of imagine all the cool things you could do, for example: you could say, “I need to provide directions to my place of business, I can say – okay, you’re going to make a left at that big oak tree,” because I can see that. And by ‘I’, I mean the application can see that, with the technology… We just – this podcast is very well timed – we just today released our national response portal, which is a SADA and Google collaboration, which is allowing for rich mapping data of the national response to COVID-19, where we can see – well not just we, this is available to the public. The public can go on and see – for every county – cases, deaths, recoveries, where to find testing sites, or places to get treatment or care, and all that’s overlaid on a Google Map, which of course is location aware, so if I pull it up on my cellphone it knows where I am and it can give me lots of pertinent information. I still think, frankly, we’re in the middle of the very early days of tapping into the power of what the collaboration between Google Cloud Platform and Google Maps can do, we’re already seeing customers do some very, very cool stuff with these two technologies. 

Max: [36.11] So Google has some pretty amazing speech technology, which includes some demonstrations that are awe-inspiring and almost like… Computers are here, and sentient, to that level. How is that being rolled out right now? How are you guys involved with development with the enterprises? What kinds of applications are you seeing driving this? I mean, the one that coems up a lot is in teh contact center space, with people doing speech detection, language detection, and prompting so that way the contact center can either interact with the caller, or maybe just prompt the agent for information that’s being discussed so the agent doesn’t have to search for it. That’s a pretty specific use case – what are you seeing enteprirses taking advantage of this, and how is that being leveraged?

Simon: [36.57] Yeah so there’s been a lot of cool stuff in that space. I think it was a couple of years ago at Google IO where they had the demonstration of the automated agent talking to a real human being, and you really can’t tell the difference, you can’t tell you’re talking to a computer. And so that was a cool demo, but what can we do with the technology in real life? And we’re just sort of getting to a point now where we have customers that are actually using this toolset. The underlying technology is called Dialogue Flow, and it’s a really powerful sort of automation, AI tool that’s offered but really it’s a platform offered by Google that allows for building these types of things. It’s been around for a little bit in the sense that you can have things like a chat bot on your website, that would be able to interact with a consumer, but it’s evolved a lot, and the umbrella now that it lives under is called CCAI – Call Center AI. It’s very similar to what you just mentioned, in the sense that it is the ability now to – at the simplest – augment and provide insights to a call center agent, and then at the most, fully optimized, it can be the call center agent, because not only do we have the speech to text, where I can hear what you’re saying and turn it into text, which can then be analysed by you know, an application, but also the opposite. I can take text, and I can speak it with a computer. So when yo ustart to put those things together – and for all your listeners I swear we didn’t talk about this beforehand, this is just good timing – but you know, we actually now have the ability to pair that technology with the mapping technology and the Streetview technology. So I might say, “Hey, I can’t find your business,” I’m calling into your business, and by calling in I’m actually connecting to a call center AI virtual agent, who knows where I am, is able to see the Streetview data between where I am and where I need to be going, and can then read me visual cues for where I need to be going. Again, “Turn left at that tree.” And so, again, this is very – I think it’s still very early in what we can do with these types of technologies, but we’re already seeing things that I think even five years ago I would have told you were pure science fiction, happening in real life for real customers. 

Max: [39.12] You linked to this a little bit at the beginning in terms of SADA’s interaction and integration with your customers. It’s more than just advisory, of like, “Oh, you should use Google Cloud and these are the things in Google Cloud you should use, like Big Query versus Big Table and why.” I mean, how deep are you going into the actual technical implementation and projects with an enterprise?

Simon: [39.34] Yeah, so I feel like my answer to a lot of this is the same, which that is kind of depends. Just like Google tools are offered on a spectrum of sophistication and control, we offer services in the same vein, and so you know, at the least involved, we do – we provide just consulting and advising services, where we are hands-off and we’re simply guiding our client on where best to go. This works really well for your digital natives or your startups, where they are very, very capable, and they grew up in this technology, and they just need to be pointed in the right direction. But then on the other extreme end of that spectrum, we are fully building ground-up environments for our customers, and in some cases we’re there managing that environment that we’ve spun up. And again, I say it’s a spectrum for a reason, because most of our customers fall somewhere in the middle of those two extremes, but we do offer services across the board. So, to bring examples to this, some of our customers we will help them build an environment, in some cases we will build the environment. Sometimes we manage that environment, and nother times the customer then amanges that environment after the fact. I will say that in every case – unless the customer explicitly does not want us to – we always try to make sure to enable the client to be self-sufficient, because we don’t want to be the gatekeepers, right? We don’t want our customers to be dependent on us. We wnat them to work with us because wer bring vbalue, but not because they have no choice. And so, even when we’re fully managing our customer environments, we like to do sessions to make sure they understand how everything works, how to administer things themselves, and then still despite knowing all that, make the choice to offload that work to a partner like us.

Max: [41.15] SADA enjoys a relationship with Google that goes beyond just being a large partner and a larg reseller. I mean, you have a common lineage in executives, and management, and technical teams. You know, for the people listening, what is that and when we say that you guys are a big, important partner to Google, what does that really mean?

Simon: [41.32] It means a few things… Yes, we do have a lot of folks on our team that came from Google at one point in their careers, and I think we also – we do a lot to make sure our culture is a match for Google as well, and so… You know, for example, I’ve been at SADA for almost nine years, and I also feel very at home when I’m at a Google office, because they feel  very similar to one another. But I think really what it means to be sort of a top-tier partner with Google and to really have this type of rtelationship is that I think we’re dependent on each other to a certain extent. I think that the service, the implementation, the migration, adoption, services that we provide for our customers… I mean, we even have technical account managers that we assign to customers long-term, indefinitely, to make sure that they’re successful; I think those things pair really well with the fact that we depend on Google to innovate, to build the best in breed technologies and tooling, and to make sure they always run as well as they possibly can. You know, I think we get to enjoy that Google will listen to us, and we get to have conversations that are – you know – bidirectional, about everything from go to markets through the detailed technical details of a product. If I was one of our customers, I think the thing that I would say that is the best part of the SADA-Google relationship, is the fact that we’re in a really unique place where – and if you talk to Simon at JPL, you know, nine years ago, I would never believe this to be a reality – but we have customers that are pushing the limits of very sophisitecated tools like BigQuery. And we have the ability to take those customer needs or additional features they’re requesting, and bring that to Google product management. I don’t mean to misrepresent that every time we make a suggestion to Google it happens – it most certainly does not, but we do get to have a seat at the table, and advocate for our customers’ needs, wants and desires, in a way that I don’t think most other third-parties get to do.

Max: [43.43] As a closing last thought here… Outside of hiring SADA of course, for somebody that’s evaluating or thinking about either – migration to Google. So, whether that’s from on-prem to Google or another cloud to Google… What would you advice be in terms of, you know, pay attention to these things, think about these things, do these things, for that to be a success?

Simon: [44.03] You know, shameless plug but we have Google Next coming soon, as the online iteration of the Google Cloud annual conference, and it’s… It’s where I point people when they want to know more about the ecosystem, because you have everything from very high-level, you know, talks about the cloud strategy at Google, all the way down to deeply technical details. And so, it just so happens that’s coming up in a few weeks, and so I think that is probably a good place for a lot of folks to look. If you’re talking about Google specifically, I also point a lot of folks to the Google Cloud website, because there is a lot of really good content there, both from a you know, marketing perspective, to understand what the products are and what they’re being advertised to be able to do, but then it’s also extremely simple to just click ‘go’, and start doing this stuff. For those who are willing and have the time, I can’t think of a better way to learn what’s out there than to roll your sleeves up and get in there and actually play around with the platform. And then I think, the obvious answer is… Give us a call! You know, like I said, we’re not here to nickel and dime customers. You don’t need to fear that if you call us up you’re going to see an invoice at the end of it, that’s not the kind of business we’re running. We like to talk to our customers about these things, and help educate them. I think that’s a big part of our job, is doing that education, and so… Yeah, shameless plug for us as well.

Max: [45.29] Simon, thank you so much for your time, it’s always a pleasure and very interesting, always.

Simon: [45.35] Thank you so much for having me.

OUTRO: [45.39] Thanks for joining the Tech Deep Dive podcast. At Clarksys we believe tech should make your life better ,searching Google is a waste of time, and the right vendor is often one you haven’t heard of before. We can help you buy the right tech for your business, visit us at Clarksys.com to schedule an intro call. 

IT disasters happen.
Only amateurs get stuck.

Schedule a Call