.png)
Overview
In this episode, Jeff Dickman (Senior Director of Cloud at e360) and Roy Douber (Principal Architect, Observability, e360) dive deep into the major announcements from Google Cloud Next ‘25. From Unified Security to cutting-edge AI agent ecosystems, this episode is packed with insight into how Google is pushing the envelope on cloud, security, and AI innovation..
Listen to the Episode:
Watch the Episode:
Topics Covered
-
Google Unified Security: How Google's multi-cloud security vision is transforming cybersecurity.
-
The Rise of AI Agents: Why Google’s AI agent ecosystems could redefine customer experiences and enterprise workflows.
-
Cloud WAN and the Borderless Backbone: How Google's global network is a game-changer for enterprise connectivity.
Key Takeaways
-
Google is leading in multi-cloud security with its Digital Twins approach, simulating environments like AWS and Azure to identify and patch vulnerabilities faster than competitors.
-
AI agents are moving beyond chatbots: Google's Agent Development Kit allows enterprises to build complex, multi-agent workflows that improve customer service, retail personalization, finance operations, and more.
-
Cloud WAN offers a 40% improvement in global network performance, enabling faster, more secure, and more resilient connectivity across regions — without the need for complex SD-WAN setups.
Read the Transcript:
*Tech Sessions Ep. 11: Google Cloud Next '25 Recap: Cloud and AI Innovation
===
Tech Sessions Ep. 11: Google Cloud Next '25 Recap - Cloud and AI
===
Roy Douber: [00:00:00] think across the board, first of all, I think Google is leading the pack from a security perspective. I think the digital twins concept, which for security is unique.
I don't think anybody else is doing that. and I think it puts them far ahead of the pack. So just to give an example of what they're doing that I think is super interesting is they're essentially taking a clone of the Google environment of the AWS environment, of the Azure environment and then it's essentially a twin and then they try to break it.
INTRO
---
Jeff Dickman: Hello everyone, and welcome to the next e360 Tech Sessions podcast. I'm *Jeff Dickman*, Principal Architect with e360 and senior director of our cloud practice. With me, I have *Roy Douber, *Roy why don't you go ahead and [00:01:00] introduce yourself.
Roy Douber: Hi everyone! *Roy Douber* here, Principal Architect here at e360 with a specialty in observability. Pleasure to e-meet everyone again.
Jeff Dickman: It's good to be back, Roy. I think it's been quite a while since you and I have done a podcast together, hasn't it?
Roy Douber: Yeah, it's been quite a bit, but this is an exciting time for another one, I think there's a lot of interesting takeaways from Google Next.
Jeff Dickman: Yeah, absolutely. So speaking of Google Next, it was about a week ago and you and I were both there we were hanging out doing the thing and at the keynote, Google did quite a few announcements, right? It was pretty heavy in AI and security and there was some underlying infrastructure announcements as well as developer and operator experiences.
Was there anything that you picked up that because you look at it differently than I do. I look at it more from an infrastructure standpoint, but you've got the application and observability aspect.
Roy Douber: Yeah. So I think there's a strong focus [00:02:00] on AI, obviously, right? Like you can't, nowadays, that's the key word.
But there were also a lot of other announcements that were very interesting, and where Google is becoming extremely strong. So I think in this conversation, what we wanted to cover, you know, is Google Unified Security. We wanted to cover some of the AI agent ecosystem and everything surrounding it.
We also wanted to cover Cloud WAN, which applies to, you know, many of our clients and many companies and enterprises out there. I think that's a very interesting solution. And, you know, we'll probably end with something along the lines of streamlining application lifecycle and every, some of the other developer announcements that happened at Google Next.
Jeff Dickman: Yeah. And there were a lot, I think I red somewhere that it was something like, over 200 [00:03:00] announcements were actually made at Next this year. Yeah. So there was quite a bit to cover, so let's talk about one that I thought was really cool, which was the Google Unified security, as you know, I have a background in AWS and Google.
And so, you know, I definitely think about things in more of a multi-cloud or a hybrid cloud standpoint. And so knowing that Google is already at the front of security with security command center and you know, other products that they offer in that area. The Google Unified Security was really intriguing to me because it now allows you to have more of a hybrid cloud standpoint from what you can look at.
It integrates with AWS, it integrates with Azure, of course it integrates with Google, but this addresses a lot of complexity in the modern enterprise environment as far as how they're viewing security, it also has a lot of relevance for composable infrastructure. And when you're deploying, being able to detect things that are going on in the environment fairly quickly.
When you look [00:04:00] at it, was there an area where you thought, wow, this really has a great use case for our customers?
Roy Douber: Yeah, so I think In our verticals and the retail side of things, I think across the board, first of all, I think Google is leading the pack from a security perspective. I think the digital twins concept, which for security is unique.
I don't think anybody else is doing that. and I think it puts them far ahead of the pack. So just to give an example of what they're doing that I think is super interesting is they're essentially taking a clone of the Google environment, of the AWS environment, of the Azure environment and then it's essentially a twin and then they try to break it.
From a security perspective. So they go in, they run [00:05:00] multiple permutations against that environment, against that read-only environment and try to break it. And when they do, they can backwards engineer how it was done and let the user know, let the company know how they were breached and how to fix it.
Very, very unique way of doing things. I don't think there's any other company that's doing it that way, and, you know, a very AI centric approach. I think in the retail space, there is, you know, obviously you've got point of sale, you've got e-comm, you've got all this infrastructure spread across the world.
So providing them that unified view, super valuable, right?
Jeff Dickman: Yeah.
Roy Douber: That's something that most retail struggles with [00:06:00] and where you see the breaches happen is usually where you least suspected sometimes these old point of sales machine that haven't gotten upgraded, that have some kind of a zero day attack happening against them.
Jeff Dickman: Yeah, I thought about healthcare right when I heard about it. You know, there's a lot going on with HIPAA and compliance and protecting PII and Google Unified Security has a lot that it can do there as far as making sure that you're securing patient data correctly and that you're doing the appropriate threat detection on your environment.
Roy Douber: And you know, finance too, from a compliance perspective, they give you that automated reporting, automated compliance, things that some of the products out there have, but not in this unified multi-cloud with the additional AI capability, I think it really separates Google from everybody else.[00:07:00]
Jeff Dickman: Yeah, yeah. You know, I'm really glad that our security team is on top of this, if you haven't, I'm gonna do a shameless plug for them here, they did a podcast earlier this week specifically talking about Google Unified Security, if you haven't had a chance to check that one out, I encourage you to go back through our history and you should be able to pull it up.
And it was a very cool podcast for them to take care of. And there are some funny pictures in there too. So Roy, I know when we were at Next and we were talking, you were super excited about the AI agents and agent space, do you wanna talk about that a little bit?
Roy Douber: Yeah. So I think, how do we say this? I think the AI announcements they come heavy and many organizations have yet to implement them. But I think that Google, again with their APIs, with their kind of cohesive approach to this, they're winning again. So they've got in one place, [00:08:00] they've got the agents, they've got all the SDKs and ADKs AI and their focus is around actions and workflows and they've built an ecosystem that your average AI practitioner can go in and basically build the next product.
One of the things that they showed us that was super interesting and kind of ties back to the media and entertainment space is they brought all the partners to the sphere and they showed us how they've built the tool set that is available to most of the world today, right? and that's through building remaking The Wizard of Oz from the old movie. They took the old movie and reproduced it so that it fits on [00:09:00] the spheres screen. And in order to build that, they had to create a bunch of capabilities that formerly did not exist, for example.
If you, if you considered the resolution of, back in the day, you had an area of the screen that you would have the characters on.
Jeff Dickman: Well, yeah, it was actually like smaller than 1024 by 768. Right. I mean, it was pretty small.
Roy Douber: Small grainy. And so not only did they have to up the resolution on every scene that existed, but they also had to.
I think they call it in painting, in paint, what it would've looked like had the screen been bigger and the camera was able to capture more on the screen.
Jeff Dickman: Yeah. It's just amazing.
Roy Douber: Basically generate frame by frame what the Wizard of Oz would've looked like if the camera was of the [00:10:00] resolutions of today's cameras, essentially.
Jeff Dickman: Yeah, and effectively for the, like, I think you could conceivably consider it like the biggest wide screen on the planet. Yeah, right. Talking about the sphere and so yeah, it's just incredible to me the way that they can extrapolate out, like when the character is off screen, what they would be doing off screen, and then as they, as you know, they're supposed to come back into the shot, they show them walking back into shot.
Yeah, and the the level of training that they had to do for the AI for each of the individual actors', mannerisms and things like that is pretty incredible as well. So that when they're off scene and they're still, you can still see them. It doesn't look like a robot or, you know, a CGI effect that they've put in to do that.
Roy Douber: Yeah, yeah. And they had to train things, frame by frame. they pulled in context from the original design of the movie, like basically they created a model where like the old notes of the directors from Wizard of Oz and from former movies of how the shots were [00:11:00] being taken. We're all introduced into this model so that they could rebuild things literally frame by frame.
And it looks incredible. Unless somebody told me I wouldn't be able to tell, basically is Yeah.
Jeff Dickman: I told my family about it and we've all agreed in August when it's released, we're taking a trip to Vegas to go see it at the sphere. So
Roy Douber: I think it's worthwhile. Yeah,
Jeff Dickman: the kid's pretty excited about that one, so
Roy Douber: And I guess the awesome thing about it. Is as they were developing this movie, they were also developing the tool set that surrounds all these new AI innovations and the AI agent ecosystem and VO which is their video, you know, in painting up resolution, basically you know, tooling around the creation of this movie is now available to all the studios [00:12:00] out there and all the people out there that want to create content in this matter.
Jeff Dickman: Yeah, I was watching a presentation on that and it was really incredible. They took a still shot of the Vegas skyline from probably, I'm gonna say it was several thousand feet up in the air. So you've got, you know, a pretty good subset of the skyline, and then they turned it into a video.
And the video pans left and the video pans right, and you can see a helicopter flying in there and you can see the cars moving on the street. And you know, it was really just incredible the amount of detail and resolution that was in that picture. And, you know, I was sitting there and I was thinking about that, and I was like, man, just a year ago ChatGPT couldn't even get like the letters and words right in pictures, you know, you'd ask it to do something and it would just have like gibberish in there.
And now, we've got this level of detail with, you know, cars and, you know, words are correct everywhere and, you know, it just looks fantastic. I mean, we live in exciting times when it comes to that kind of stuff.
Roy Douber: Yeah. It's, so, it's amazing. I'll give one more [00:13:00] example because I think it's just so super cool in the U2 sphere experience. You can go to the virtual one now too, they deconstruct the city of Vegas. So basically it starts with Vegas as it is today. And slowly but surely, all the buildings for one of the songs, all the buildings go away and you end up with just desert. Again, I believe that was all kind of like AI generated at least partially AI generated, It took all the pictures, dropped it into probably the same model they used to create the Wizard of Oz and slowly deconstructed Las Vegas, as a part of the song
Jeff Dickman: Did you get a chance to look at the agent development kit or the agent to agent protocols?
Roy Douber: Yeah, so I think it's really interesting and again, I think this is early days for this [00:14:00] tech. But I think it's impressive, meaning I was there for the initial renditions of this tech. Like you could say, okay, I have a project manager, I have an engineer, I have an architect, I have a storage engineer, and I want to build a project plan to do X, Y, and Z.
And it would always, all the agents would confer with each other and you would get a pretty decent result back in the day. But they've taken it a step further where now you can develop applications around this and have a bunch of checks and balances, so you can create workflows and actions at various steps in the pipeline, essentially.
And allow you to verify tests, the output modify the output at certain pieces of the pipeline just like you [00:15:00] would in a real world project.
Jeff Dickman: Yeah,
Roy Douber: I think this is an area that's slated to grow, and I believe that over time we're gonna see. A lot of the chat bot use cases has moved to this kind of technology where it's more thought through and we act rather than just a question and answer interaction with AI, you now have, call it various AI agents that are working on your behalf to achieve certain results
Jeff Dickman: Yeah, well, when you even think about it right now with the agent development kit, You can have your agent interact with services outside of Google, or also services inside of Google, of course, but you know, they did a really great demo of a voice agent that was able to interact with the user shopping cart and, you know, make a phone call and escalate to a CRM for approval to give a discount.
And so [00:16:00] you know, retail, I think there's a ton of potential here, within retail for the personal shopping assistance and you know, automated inventory management. And you know, that really great customer service where you can have the AI handling the mundane conversations, you know, with customers.
And you can have your customer service team focused on the higher order questions and the higher order problems, you know, that the AI escalates to them and then they resolve it, and the AI provides that resolution back to the customer, in a way that's friendly and, you know, handles things and you know, even remembers details.
Like I thought it was really funny in the demo, the the person doing the demo was talking about a pickleball tournament, on the call he told the agent that he had a pickleball tournament on the day that they wanted to do the work. And at the end of the call, as the agent was hanging up, it's said, good luck on your pickleball tournament.
Roy Douber: Yeah.
Jeff Dickman: Right. So, I mean, it's just fun stuff like that I think, you know, that's a very creative way of having the AI out there and making it a little bit more [00:17:00] personable. And also it improves customer perception at the same time, right? if you've got something that remembers contextually that you said things and then is able to say something at the end of the call or ask you a relevant question later in the call that to me or details that customers remember.
And it says that you've gotta focus on the customer experience as you're, you know, trying to sell them more product.
Roy Douber: Yeah. I remember the Pati example, right? Like one of the guys was buying fertilizer for his yard, and the AI agent was interacting and it noticed that the fertilizer that they were buying was not the perfect fit.
It was good enough, but then it made a recommendation and said, Hey, by the way, I'm seeing you've got this for your petunias. I think this fertilizer and this, you know, it gave them a, a couple of product recommendations on the fly would do better and make your petunias grow at a better rate, it gave them all the reasons why.
You should basically [00:18:00] swap the fertilizer and it's all, and it's the same costs and I thought that was really, really cool. I think you know, it takes away a lot of the research that you would have to do or sometimes you don't even have time to do. Right.
Jeff Dickman: Yeah.
Roy Douber: In real time as a part of your shopping cart experience.
Yeah, think about finance. I mean, just on the finance side of things, you know, one of the things I hate doing is calling my bank, you know, because it's, press one for this, press two for that. You know, verify your birthday via, you know, you talk to it or and it's voice recognition, but it's not like thinking it through, but, you know, having a AI agent that, you know, they have, they have a ton of data on me, they know a lot of stuff about me, and having that agent check in with me on different things, you know, when I'm just calling for something, you know, there's a lot of opportunity there to provide a higher level of customer service which could lead to retention as well as you know, upsell services, right?
Like, hey, we noticed that your your [00:19:00] CD that you have in here is expiring in a month. Do you wanna set that up to automatically renew or do you wanna close out the CD? You know, there's all kinds of options there, but you know, more context that they have on me, the better those can be.
And then having it be conversational. Instead of press one, if you want to do this, press two. If you wanna do that, to me that's way more valuable and it leaves me hanging up the phone thinking, wow, I feel like I got the service I wanted instead of just frustrated because I had to navigate a menu 12 levels deep.
Yeah, it's really interesting right, like all of these use cases coming together, you've got the model context protocol coming out of philanthropic labs, which is slightly different than the A two A protocol that Google develops or agent to agent the focus of the Google side of things handling agent to agent secure communication and collaboration between the agent [00:20:00] while MCP, the model contact protocol offers the infrastructure and operations layer interacting with various APIs out there and they're actually complimentary good. So I feel like Google, you know, there's a lot of content out there that suggests that Google is competing with philanthropic, where really I see those two tools as complimentary goods, right?
So you can build all these kinds of use cases that add additional context. And there's also this agent to agent protocol that allows you to ensure that not only, are the agent communicating quickly, so basically in their own protocol that maybe doesn't require human language, right? Like they're trying to speed things up, but also you have access to now thousands of APIs, [00:21:00] CRMs, you know, ERPs, all these tools that are cumbersome to deal with where all the data is stored, are now very, becoming more and more accessible.
Jeff Dickman: Yeah,
Roy Douber: And one of the demos that I saw and in the dev session highly suggest if you're a developer, check that out, was they're adding a data agent which basically allows any person to go as long as they have connectivity to a set of data or a database or you know, it could be a NoSQL database or a se a relational database, transactional database, what have you, you can ask it questions about the data and ask it to manipulate the data and join data across various other data agents. So the possibilities for reporting, building sales reports, building forecasts [00:22:00] is now open to almost everybody. Okay. And, you know, with a caveat, obviously AI, I would say is 90 to 95% accurate, right, and improving. Generally you still need some for the complex use case, you still need some knowledgeable hands but I do think there is incredible opportunity
Jeff Dickman: So when you say you can just point it at the data, right?
Roy Douber: Yeah.
Jeff Dickman: I have worked as a data analyst you know, in years gone by and so I'm familiar with like getting a database handed to me and having no idea, like what I'm looking at and having to go through and build a metadata table to like give me the context and allow me to then connect it into the BI tools that allow that to happen.
This agent is able to do that function where it looks at the data and it's able to make the connections and that metadata, if you will, that points back to other data and allows you to make the linkages and things like that's sort of what you're describing [00:23:00] here, right?
Roy Douber: Yep. that's where it's headed you know, again, it's early days and the technology is relatively new, but you know, what they were demoing, is it building out reports, understanding the data, scanning the data, looking at all the tables, you know, adding context to the data as needed, again, if you have sales data in there, deriving, you know, which customers are likely to merging, merging all the data sets and giving you a likelihood that the customer will be interested in the solution that you're trying to provide.
Things that would take even the more advanced data engineers and data scientists. Long time to do.
Jeff Dickman: Yeah.
Roy Douber: It leads you in that direction, so one of the things that I think is really interesting in AI age is that you're now able to take on projects that you didn't think were [00:24:00] feasible before.
Jeff Dickman: Well, that's really one of the big things that AI has done right, is it's democratized data and the ability to access data and to make, you know, informed decisions based on data instead of making informed decisions based on your gut feel. And you know, I remember years ago, it was one of the frustrations of the quality manager that I worked for, that he had to wait for me to run the reports and he would tell me what he was looking for, and then it would be, you know, anywhere depending on like, which systems I had to access to get it would be anywhere from like 20 minutes to days.
to get him to report with the data that he wanted. Then he'd be like, no, that doesn't look right, and then it was back to the drawing board. Now, I mean, I've done some of this just with some data that we have, and I'm able to run reports in minutes, if not, if not faster. That really just presents exactly what I'm looking for.
And if I'm like, no, no, that's not right. I just change my plain language query. And then I'm able to update the, you know, the data that I'm seeing in real time. So the democratization of data is huge to me as far as like the capabilities that AI presents for businesses because people [00:25:00] can't afford data scientists.
Not everybody can. Right? Some people can, but I most companies can't. And so the ability to take an agent and say, I'm gonna have an agent that's going to do my engineering for me, you know, like cleaning up the data and getting the formatting right, and then I'm gonna have another agent that's actually gonna do the data science for me.
And you know, if it gets to be super complex, you might need to hire a data scientist, but for most organizations, I think between those two agents, you would be very able to function and get the reporting that you needed out of your data.
Roy Douber: Right, yeah. And I think it just increases the velocity and I think there's a lot of ideating going on now that have happened before because people now have the time in the day to kind of ideate and they, you know, there is that added helper that allows you to accomplish things that may have even seemed like it wouldn't be worth the effort in the past. Right? Like now I can [00:26:00] spend some time building my time sheet automation tool that I've been working on because I think there's merit and with AI, it makes things go a little faster and I can increase my velocity as a developer as an example.
Jeff Dickman: Yeah speaking of faster, did you see the announcement for cloud win?
Roy Douber: Yeah, I saw that. I think that's another interesting announcement, right? All those Google undersea cables, the entire Google network not exposed to the world. I think again, another key differentiator for Google.
Jeff Dickman: Yeah, 2 million miles, that's what they said, and so like when you're talking about a large organization enterprise that's working with AI and that they probably have, you know, data in multiple different locations. The the 40% improvement is a big deal, right? [00:27:00] Like that's more than I think most people would expect. Like if I have an MPLS network or, you know, something along those lines, to see a 40% improvement in latency across that is just like, I mean, that's a network engineer's dream, you know, to see that kind of improvement. And so when they were announcing that they're making that available, that they're making the backbone available to enterprises.
it's like the opportunities to be able to bring different pieces of the organization and data that might be in Europe and the United States together to get meaningful insights out of it, especially when you're talking about a global economy, right? that to me is just like really important even inside the United States, right?
If still talking about like retail in the economy to have those point of sale transactions happen more quickly. I mean, most of the companies that do point of sale transactions, they have a latency SLA between the customer and the point of sale device and [00:28:00] them as well as back into the processing environment they have, you know, there's SLAs across the board for that stuff.
And so to be able to now reduce that and, you know, make achieving your SLA little bit easier, I think that's gonna be kind of a big deal for a lot of these companies.
Roy Douber: Yeah, I think there's also a security angle, there's all these legacy SD WAN implementations that businesses have that can go away.
It's just such added complexity, so many network engineers that have to manage it, and I think, you're gonna need less of those.
Jeff Dickman: Yeah.
Roy Douber: Yeah, you go with something like this. It's a managed solution, it works at Google Scale, it runs their various microservices across the world, it runs you know, everything that we touch that is Google, including search already and probably for many years.
And they've just obfuscated all the complexity away.
Jeff Dickman: Yeah.
Roy Douber: And now [00:29:00] you're able to use it, I think it's brilliant and in years past, having worked on the network side and understand all the complexity and intricacies and downtime and when things go wrong, I think there's a lot of thinking that needs to be done at the organizational layer when products like this come around because it just lessens all the complexity it's a global backbone, it's low latency, it's high availability, everything that you've dreamed of that is now managed by tens, if not a hundred, depending on the size of the enterprise, you know, network engineers, you can do away with.
Jeff Dickman: Years ago, I worked for a company, I worked out of one of their remote offices in Fresno and their main headquarters was here in Colorado. And we would have to [00:30:00] copy files from Fresno to Colorado and, you know, like we're talking, hundreds of megabytes up to gigabytes, this was a while ago, that was a lot.
And it would take hours to move that data and so as they were talking about this, I was thinking about, okay, Google's backbone is huge, right? So if you have a media company that's moving files, let's say it's between their maybe it's their offshore rendering farms or something like that.
And then they're moving them back into the states that being able to move those files across Google's backbone would allow them to have a much faster return on, Hey, okay, this is the file, let's look at it, not good, okay, let's do it again. Instead of waiting hours to, you know, maybe days for these you know, hundreds of gigabyte files to copy over some other network mechanisms.
So I think there's a lot of potential for that as well as moving, you know, high resolution healthcare files from, you know, you've got somebody in California with cancer, right? and the doctor that knows the most about that is in New [00:31:00] York.
So being able to port that file across Google's backbone would get it to that doctor faster. You know, the imaging files and you know, the scans and all of those things. And so I think there's, you know, a lot of benefits there as far as what people can see by leveraging Google's backbone.
Roy Douber: Yeah, I mean, to add to that, right, think of like just your network segmentation and policy enforcement across the globe. I mean, just imagine, right, like now you've got one network and that's it. There is no separation globally, you don't have to have a different rule for India, a different rule for the US, right, like you've got it all planned out and you can come up with a common set of policies that are enforced globally, the security implications there are, you know the organization is more safe.
Jeff Dickman: Yeah. it integrate with the Next Gen firewall. So you can have your policies in the Next Gen firewall and VPCs are global, [00:32:00] so, you know, it's gonna make things very interesting as far as setting up infrastructure and getting customers, networking, going the way that they want it to work.
Okay, so I know this one is very dear to your heart, Roy. This is the the streamlining of the applications and everything that's going on in that area with Google, Google has a push for an application centric cloud. Do you want to talk about that and help me understand what that is?
Roy Douber: Yeah, so I think just in general what we're seeing with Google is they are consolidating their offerings into centralized offerings. We've seen it with GUS, right? So Global, unified Security across the board, they're doing the same with their observability offering. They're making everything application centric. They're looking at the management of the application as a whole and not just the underlying [00:33:00] parts.
Right? Like, oh, yeah, you know, I have a VM, a database and I think that the end goal here is to make things easier for the developers and the operators, which I think is at the core of the whole SRE movement, right? Like, no less toil, get me the data I need, when I need it, where I need it, and so they're making a compelling offering that competes with some of the companies out there that offer up.
Observability and security you know, as a service so, and I think they have the features that in some cases they're winning. In some cases they might not be winning, but it's compelling enough to say, you know what? I don't need to pay the premium price. I could just go with the tool set that Google presents to me.
Jeff Dickman: Yeah.
Roy Douber: And it's more than good enough. It's something that I wouldn't have been able [00:34:00] to even conceive of a couple of years ago, right
Jeff Dickman: right.
Roy Douber: So, yeah. One of the tools that I found super interesting was the application design center, you're an architect, I'm an architect. You know that one of the most cumbersome things is templates and blueprinting and
Jeff Dickman: oh yeah.
Roy Douber: Bringing an idea to life, right. And bringing a designs, whether it be software, whether it be infrastructure to life. And Google has created a tool that is currently, I think it's in preview, but I they promised it would go GA fairly soon, where you prompt and it creates the design within the Google context for you, inclusive of the diagram, inclusive of the Terraform code.
And you can ask it to do even more than that. So with built in best practices, so [00:35:00] you're well architected follows well architected principles, very, very interesting. Very, very powerful and again, such a clever use case of AI. Right?
Jeff Dickman: Yeah, so this to me it sounds like, a couple years ago, Google had a service that was, I think it was called the Architecture Design Center. And it was sort of a place that you could go and they had like the diagram board and you could draw your little infrastructure pieces on there, and then it would write the Terraform for that.
They took it down a while ago and they said that something new would be coming that would replace it. It sounds to me like this application design center may be like the first step in replacing that with something that can do not just the application, right? The cloud run and you know, the GKE containers and everything else that you're gonna have running there.
But it'll actually [00:36:00] allow you to deploy the infrastructure as code as well. So we're really talking about composable infrastructure with the application on top.
Roy Douber: Yep, and they've introduced somewhat of a rag construct here as well, where you can create spaces and projects for your own team. So say your team has a set of best practices that you wanna inject into the process, you can and it will try to follow based on your prompts the best practices that you've devised for your internal team.
Some of the qualities and the features that you would want to inject into your application development and it does a fairly good job based on some of my tests.
Jeff Dickman: So somebody's finally gonna follow that style guide that we wrote, right?
Roy Douber: Yeah.
that's gonna follow the style guide but also, you know, follows, let's say you need a certain compliance, you know, certain [00:37:00] pieces of Google are compliance that certain may not be. So you can inject that opinion into this design center for each team, right? So each team has their own biases and things that they'd like injected.
And it allows you to do that and it's built into the tool, right? So like each team gets their workspace, they operate within their workspace. And it seems to me like there's a learning that happens behind the scenes. As you develop more and more applications and deploy more and more application, and this takes it all the way to deployment.
So not just generation, right?
Jeff Dickman: Yeah.
Roy Douber: You're also deploying using this
Jeff Dickman: So after deployment you've gotta monitor it, right? application monitoring which is in preview seems to cover that. How does it compare to other application monitoring solutions that you've used in the past?
Roy Douber: Again, I [00:38:00] think it's the push towards consolidation of everything in one place. You're seeing other observability players doing it, you wouldn't necessarily expect the hyperscalers to do it as well but they are I think Amazon does well. I think Google does even better and there is a massive focus on this space right now and you can now do everything in context, application performance, health, user experience linking back to applications defined in the application hub, which is another service that they've built, which is basically like a central registry for operational views of all your applications.
So again, like if you think of like your Dynatrace or your Datadog of the world if you ingest all that data via [00:39:00] agent into the, and then it creates those centralized views for you where you could look at the application, the components, the services and that's their new application hub. So it essentially indexes and creates a catalog of all your applications and then you, in the application monitoring side, that's linked back up to that app hub.
So various different applications that allow you to focus on unified views.
Jeff Dickman: So, while application monitoring allows you to look at the performance of your application as a whole does it also allow you to trigger actions against the infrastructure, like perhaps a scaling event or expand that the number of nodes you have in a cluster or anything like that?
Roy Douber: Yeah, and that's again, those were features that existed in the old, but now you've got everything kind of, how do you say this, logically flowing, right? Like I [00:40:00] have my app catalog, I see that my application is red I click into it, it points me to my application health or the user experience of that application.
And everything kind of logically and cleanly flows together. And then you can go to the cloud hub, which is another view that they've created in the observability space, in this consolidated space that's focusing on infrastructure, resources, supporting, again, the applications defined in the app hub.
So you've got the application piece, the infrastructure piece, you've got your app catalog at the top, you've got a way to design and deploy and push the applications to production and so again, I think the ecosystem is very compelling and especially, you know I think Google is far ahead in this space actually.
I [00:41:00] think they're making deploying applications very simplistic. Managing those applications is now becoming more and more simplistic, and I think it jives with the whole AI motion.
Jeff Dickman: Yeah. I think that what I see is Google is really creating this whole chain of, you know, operational efficiency, application reliability, and allowing you to put those into a secure operations framework.
Which is something that doesn't really, you have to build that in other places, but now you just have to sort of configure it with Google and you can get the results that you want out of Google Cloud. So it's all very cool. Man, there was so much, so much that was announced at next this year.
[00:42:00]
And I know we're about a week past it, and you and I have had a lot of time to sort of geek out on it and watch the recaps and do all those things. And you know, even after doing all that, I still thought, man, these were the four areas that I really thought, you know, they're game changers for a lot of organizations.
You know, the AI options that they now have with Gemini 2.5 and you know, the vertex AI studio, and the ability to put all of those things together and to just create, you know, bring your ideas to life, you know it's not something that you have to hire like legions of people to do anymore.
You can talk to the agents and you can make it happen. And then, you know, when we talk about cloud WAN and you know, the performance improvements that's going to bring to organizations for having their agents connect to their data as well as how they move data around. It's gonna be very shifting.
Especially when you think about the gravity that data has and you know that gravity is not quite as significant now as it, as it potentially was.
Roy Douber: With all, with all these tool sets, I think you can. I mean, to tie it all together, you can design a well architected application that runs reliably over a global network that they expose to you.
[00:43:00]
Jeff Dickman: Yeah.
Roy Douber: Secured, end to end by a unified solution that leverages AI and that is monitored and looked at holistically. It's a fully integrated vision that I don't think is being executed as well anywhere else.
Jeff Dickman: Yeah,
Roy Douber: and I think that's kind of my, like if you had to put it all together into one place for people to consume, that's what I would say is like all of these things are coming together and making the world easier for anybody from a developer all the way through operations, right?
Jeff Dickman: Yeah.
Roy Douber: and soon, I don't know that you're gonna be able to tell much of a difference between all of these people. I think they're just to be operating within a space that's created to build the next set of applications that are AI driven.
[00:44:00]
Jeff Dickman: Even going as far as like the idea guy, right? You know, that guy that's got all the great ideas, but nobody to implement them, they're gonna now have, you know, a team of people through the agents that can implement these things for, I mean, the opportunities that, you know, become available for folks just like everyday people to take advantage of these and bring their ideas to life is just, I mean, it's just incredible when you think about it.
So Roy, thanks. I appreciate you taking the time to chat with me today. And you know, it's been always, it's always fun to talk to you and to geek out on the announcements that the cloud providers are doing and it was fun hanging out at next with you and picking your brain while we were there as well.
[00:45:00]
So again, this was the e360 Tech Sessions podcast, we we encourage you to to take a look at how these announcements apply to you we'll provide some URLs in the, in the comments as far as where these announcements are located, and then if you have any questions or anything or want to dive deeper into any of these technologies, you feel free to reach out to e360, Roy and I are happy to chat about them, dive into infrastructure or the applications or observability or you know, anything in between those pieces of it, our website is e360.com and I'm *Jeff Dickman *and I thank you for listening to our session.
Roy Douber: Thanks everyone.
