Ep.168- Google Next Recap: Where is Google going with Gemini and Agentic AI with Koby Phillips
Subscribe to the Next Level BizTech podcast, so you don’t miss an episode!
Amazon Music | Apple Podcasts | Listen on Spotify | Watch on YouTube
Transcript is auto-generated.
Welcome to the podcast designed to fuel your success in selling technology solutions. I’m your host, Josh Lepresto, SVP of sales engineering at Tilleris and this is Next Level Biz Tech.
Everybody welcome back. We got a special episode for you. We have been out there in the market, lots of industry events, lots of things happening, really pay attention and paying attention to what’s going on and what that’s going to mean for us in the partner community. So today’s episode is titled the Google Next recap. Lots of good stuff happened a couple of weeks ago. I went out to Google Next this year.
know, Kobe and I, talked about this, you know, late last year we went out to AWS re-invent and so thought it’d be awesome to have Kobe Phillips back on VP of the cloud practice to flip the tables, interview me and we talk about Google next. So Kobe, welcome back on as not only the only person to ever host, but the only person the second time to ever host. So good to you back, man. Well, I’m certainly honored. I appreciate you having the confidence in me to take the reins for an episode or two and happy to be here.
Awesome, man. I’m going to turn it over to you. Let’s talk through. I know you’ve got some good thoughts here and excited to tell the partners about what we saw and what it means. Yeah. mean, for those that couldn’t attend, Josh, kind of just give the landscape of what Google Next is. I know we had the previous podcast that talked about AWS reInvent, maybe any key differences between the two conferences or any highlights that really stuck out to you.
Yeah, so Google Next is essentially think Google Cloud’s annual conference. So, you know, we talked about reinvent, re-invent’s going on for years and years and years that happens out in Las Vegas. And now Google Next has been happening for the last few years. So Google Next, you know, on the surface, it might feel like, ah, this is a way for them to just show all their products. You might want to be dismissive of it. And it’s, you know, it’s just a big marketing thing. And the reality is it couldn’t be more far from that.
And so if you think about just for a second, flashback to what stands out to me is I remember five to 10 years ago when we were looking at Google Cloud and people were going, are they really gonna invest in this or are just gonna give up? And now seeing this thing, it’s a $50 billion run rate between all the products that are there, between where Google Cloud has kind of figured this out, between Thomas running that whole organization. so we’ll dive into a lot of this, but just seeing that
This is real. This is, it was very well orchestrated. was 30,000 of myself and my closest friends and a lot of good sessions, a lot of good keynotes and seeing what’s to come. And Google is really here to play and they are not, they are not a joke. Well, and I appreciate that background because if I’m sitting as an advisor and I’m, I’m looking at it just from the lens of our community, I’m like, what’s the point of going to sound like Google’s not directly in.
but they’re influencing a ton, right? And they have a ton of, lot of partnerships with suppliers that are in their ecosystem that are gonna be impacted by some of these technology advancements and updates and things like that. It is a mix of end users, their partners that are, know, deemed suppliers in our neck of the woods. But what are some things that really stuck out to you as far as some of those like innovative things that we come to expect from Google? Yeah, I think the other way to probably I should frame this up before I start is
When we go out and see these things, when we go to Dreamforce for Salesforce or AWS for reInvent or now Google Next, what tends to happen is the enterprise customers start to adopt these things and these trends and these products first. And so often what you’ll see is, you know, the XYZ session that I go to, they might have, you know, a Google engineer or a Google head of product.
mixed with maybe somebody from the customer side, obviously, to do the testimonial and to talk about how they’ve already built it, it already works. And so we tend to then see those things over the next 12 to 18 months transcend into organizations of all shapes and sizes. so we had lot of, finally at a spot where we have overlap, seeing suppliers, they’re Dialpad and Equinix and Five9 and RapidScale and all these. So seeing that connection into our ecosystem and then finally seeing
you know, picking my sessions, know, 30,000 was a lot more manageable than it was at AWS where it was 100,000 people spread across eight or nine different hotels. That was a lot. This was, this felt manageable. It felt like you could get to all the things that you wanted to get to set your schedule accordingly, make your way through the expo hall. Um, but, but no surprise AI at the forefront. And so almost every session that I was in either some
some product around AI, agentic AI, and how people are already using it. Some of the big standards coming out that we’ll talk to that we, you we were just talking last week at our big Salt Lake AI Summit about the agent to agent protocol and kind of what that new standard is and how I think you’re going to see that through everything. Some security things that we can talk about that we’ve seen, you know, generally in that infrastructure and security side and, you know, really the customer experience side.
model building and things like that were some of the highlights. We talked about kind of being overwhelming for, you know, just how much it is at the AWS. And I would argue this is probably if you’re an end user, like you’re a customer, it can be equally overwhelming. analogy that I that kind of pops in my head is when you go to a home show, right. And you’re like, Oh, I my landscaping to look like that. Or I want this or that, but I don’t have the skillset to do it. And this is where the advisor advantage comes in with the suppliers that Josh had mentioned and others that that were also in attendance.
they’re the ones that can go and make all of these things happen for customers. And yes, it starts at the enterprise level, but you get a lot of that mid-market smaller enterprise that doesn’t have the skillset and resources on staff, but they see all of these advantages. So Josh, you know, you talk about AI and some of the stuff that really pops out. What in particular that stood out to you that was one of those like, wow, that’s really cool and impactful. Maybe just one or two of those things that you saw there. Yeah, I think
probably the most interesting or most maybe surprising moment was to see how some of these, going through some of the migration talks and how some of these big Google migrations have already happened. Some of these AI implementations have already happened and people are already using them. Where you think, okay, we’ve only been at this AI thing the last one to two years, how mature can…
products really already be and you’re constantly surprised You know, we talked about this again last week, but you’re surprised to see how far these things really are So I’ll talk about maybe the first one in seeing where a big OEM and the Google engineer that leads out the agent to agent Protocol and so going down this agentic AI theme and that was the theme that that clearly was everywhere but You know seeing first if I the customer
have maybe Salesforce as my CRM and I’m going down the road of Google Cloud and maybe I want to build my own models. I want to train my own models. I want to interact with other models and things like that. Well, I need, I need to make it, if I’m Google, I need to make it easy to embrace whatever a customer already has, whatever software and SaaS and infrastructure and tooling that a customer already has. Maybe they have Atlassian, maybe they’re doing Jira for ticketing and things like that.
Well, if Google’s smart, they figure out a way to interrupt with everybody. And that’s what this agent to agent protocol standard does, is it says, okay, you’re gonna use Salesforce, you’re gonna use Google Vertex for Google Cloud, and then maybe you’re gonna use Atlassian and Jira for some of your ticketing and management. Well, you’re gonna kick off a request maybe that comes in, maybe everybody’s used to using your software at the customer side, Jira, for ticket management. Well, maybe that ticket management needs to kick off.
an agent, an AI agent in Salesforce to do something and to respond to a customer. Maybe it needs to tick off something in Vertex AI because you’ve built out a small custom model around, you know, just that makes most the most sense for your business. And so I think the idea when AI first came out, it was, my gosh, how are I going to have this model over here and this AI tool, this AI tool. But now we have a standard to say, okay, here’s what state each of these agents will be in. Here’s the task each of them.
was assigned to do. Here’s the security authorization that each of them has to do. And here’s how long each of those are going to go on, right? So that now Atlassian, and if that’s the entry point where people come in and start to kick off some of these agents, they now, that Atlassian kind of brain, think of it as the brain, can manage all these other agent cards. And that was really seeing that in action, seeing something come in, kick off a ticket in Salesforce, kick off something in Google.
and have it all work together to then automatically, you know, agentically, we’re going to overuse that. seeing that. It’s not going to get overused because I mean, it’s that impactful, right? see it happen and then feedback a response to a customer via chat. See it all happen real time and because of that protocol and things like that. So that protocol was that’s a big one. I they’ve already got a hundred, a hundred OEMs interrupt on that to be able to say this is the language we’re going to agree upon.
And just, you know, let’s take a quick pause for everybody listening. Agentic AI, if you haven’t heard of it by this time of this recording, you will a lot, trust us. That will be a phrase going forward. What Josh just laid out was essentially a good example of what that is. It’s multiple applications working together in an agent to agent like status and then driving an outcome, right?
where usually you’d have to build in code and do all of these things in between. It’s now starting to work together driven by the AI protocols that are set in place and the parameters. So this is the next evolution where you go, if you look at the way AI has been, it’s been a really good executive assistant, but it’s been that employee that you had to go tell what to do every single step of the way, or at least give them a pretty good amount of guidance. This is now starting to take and be proactive in a lot of these,
examples where it’s going to go and do it for you based on how you set it up. So imagine given, you know, somebody a little bit of an idea of like, hey, go clean and organize that room, and it’ll go and do it for you versus go and mop the room or sweep it mop it, put this here, organize out there, store these here, do this, that those are the differences. So like, it’ll go and figure it out on its own now a little bit based on all of all of the directions you give it.
but you’re not having to give it every single prompt and direction along the way. And that’s just one of the many different things that agentic AI is gonna bring to the table. It’s gonna scale way past that example. But that is not a surprise to me that that was at the forefront of a lot of the conversation. That agentic AI will be the digital transformation of the coming months and potentially further in my opinion. And as advisors,
you know, hitting on that. So Josh, you hit on some really good examples of like how big of a conversation this was at Google. You know, everything’s going to pop in and go, you know, what else, you know, what about the security angle? Like, how does that work? And how does that play in? What, what did Google Next offer when it comes to like security and infrastructure conversations? Yeah. So from a security side, I’m forgetting the title of the session, but I got to sit with a guy that’s kind of essentially over
threats and sock for Google for the last 15 plus years, right? So this guy has been there for quite a long time, super innovative, but also has had to manage this at massive, massive scale. And there was a common phrase that I like, that I would love to see us start using is, I think there’s a stigma sometimes with AI of just because AI is in the conversation means it has to replace something. And I think that’s just wrong.
And so they use some terminology that I love called human in the loop. So there’s always a human in the loop because we are nowhere quite yet near a spot where we can say we have implicit trust that this AI agent or this AI software that I put in place is going to do everything that I needed to do. Even in those previous conversations where agentic AI, right? Each of those things says, if you struggle or if you get to a certain point, figure out quickly if we need to have a human in the loop.
to invoke a human agent and take that to a live conversation, right? We don’t want people to struggle and have a bad taste in their mouth from an AI perspective if we can avoid it. And so from a security perspective, it was where are we on full manual, semi-autonomous, and then fully autonomous? And we are not in that fully autonomous bucket yet. We’re somewhere in this kind of old school manual meets semi-autonomous way. And so the biggest thing that I see that they showed is one,
you know, not having to write run books anymore from scratch, not having to write rule sets, custom rule sets in your SOC, in your SIM, in those in your MDR platform anymore, right? Really being able to lean on the agent to write those. And then also just from an advanced threat hunting perspective, if you think about the positive spin of what AI can do from a threat hunting perspective, we’ve always, in security, it seems like we’ve always been in this spot where we’re always trying to
catch up to the thing that just came out, right? And find a way to get these zero days, okay, we need 24 hours, 12 hours to put something out that mitigates this threat or this gap or this patch problem or whatever it might be. So if you think about advanced threat hunting, the vision on it was being able to have agents that are out doing advanced threat hunting, crawling on the dark web and bringing those back.
and creating rule sets before you even have to think about creating a custom rule on something that you didn’t even know existed yet. And so that was a big thing from an advanced threat hunting perspective. So more proactive management again versus reactive response on the non-security side. And what I’m kind of hearing is everything’s built with security framework in mind, From every lens of technology nowadays, Google’s obviously no different here. What, you know,
What in particular in the infrastructure piece did you hear and see from the event that maybe stuck out to you as well? You know, a couple things. There were some similar trends that we saw, like we saw at reInvent, you know, coming out with the new chips, the new TPU chips that are better at inference, know, version six or whatever it was, right, can handle more, more teraflops, more inference models, more parallel processing, all these things that
models need to train. And so there’s clearly a little bit of a theme here of, okay, if you’re going to come into our ecosystem, we have to give you good hardware on our underlying equipment, if we want to leverage our infrastructure to train on our models, right. So the other piece of that, you know, in addition to obviously kind of Gemini being that forefront, and we’ll come back to that is really Vertex AI seems to be kind of this core platform.
that is the standard of all model building and all things that go in and out of model building, both from a security perspective, from where in the infrastructure I need to tie this into. Is that a database? Is that a BigQuery thing? Is that a Firestore thing? You know, a NoSQL database. And so seeing kind of interop between all of those and how Vertex seems to be at the epicenter of that, or maybe it’s just…
you know, storing it somewhere else besides, you know, Firestore, Firebase, those kind of things. Infrastructure and database and security, were, all three were kind of always mixed into the Gemini and Vertex conversation. Yeah, I’ll come back to the Gemini because I do have a couple of questions on Gemini in particular. But again, just to kind of hammer that home, it’s what I’m loving about, you know, when we went to re-invent what I’m seeing from Google and how that’ll trickle into our ecosystem through our supplier partners.
and their ability to deliver on behalf of their customers, Is the welcome, we can work with that mentality. Oh, you have this? Yeah, we can work with that. You have that, we can work with that. And that hasn’t always been the case in technology. There’s been a lot of like, ooh, we don’t work with that one. And we don’t work with that one. And that’s becoming less and less the case, especially in these hyperscalers. And they’re allowing for that. And then just in a competitive landscape, what ends up happening after the fact, right?
is the ability to then go from hyperscale to private cloud and then see what they model in after it to stay competitive as well. So it’s a really cool like kicking off point and everyone’s really excited about it. And so that is, that’s what I’m getting really, really excited about. you know, what Google’s really pumping kind of going back to the Gemini piece of it is, you know, where it’s at in the landscape.
as far as their AI assistant and things like that in comparison. And that’s their go-to-market product against, Microsoft Co-Pilot or even Bedrock from AWS. Do they pump any new features coming out of that? Yeah, I think kind of this constant compete of, seems like whatever we see in ChatGPT, whatever we see in AWS Bedrock, I’m going to see Bedrock as much in this conversation, but when we see what the latest and greatest
this week in the amount of parameters or the model or the reasoning versus, you know, standard model that we have in Copilot and Gemini and then, you know, over here OpenAI seems like a constant every other week. Oh, okay, this one added deep research now. So now besides just getting quick answers, Google and Gemini added deep research, which means I can go pull from a ton of sources, take a little more time, but pull from a ton of sources. So if you haven’t used
the same way maybe you’ve used Chad GPT and OpenAI just for your day-to-day kind of questions and answers, try out Google Gemini and hit the deep research button and watch what it’s able to do to help you pull together a really comprehensive analysis around a complicated question. And so seeing that really starting to expand multimodal, so being able to do voice and video and image and things like that, they’re catching up a little bit on that.
that piece of it as well. And you you saw things like, if anybody hasn’t played with this, it was really cool, the Notebook LLM, where you give it a little bit of information about yourself, and then it goes and creates, this is powered by Gemini, it goes and creates a podcast, essentially. imagine, imagine I fed it a bunch of information about you or a bunch of information about me, it now then goes and creates an audio file that’s a podcast, you know, where somebody’s interviewing information or interviewing things about one of us, talking about us. So it’s just wild.
contextually, how it’s getting it more and more and more accurate. What’s really interesting is like, you know, I see a lot of the stuff that’s out now just as like, oh, that’d be a really cool idea, but there’s no reason to go and build it like at all. But now it’s so easy to build and put out there that you get that type of thing. You’re like, hey, wouldn’t it be cool if it seemed like somebody was talking about it on a podcast? And they’re like, yeah, but who would spend time to go do that? What’s the market for it? Other than just being like a cool idea? Yeah. Well.
there is a market for it, it is a cool idea, and now they’re delivering it. And we see more of that. You never know what’s gonna add value to somebody’s day. Well, and the thing that I would call out is I’ve always been a critic that I didn’t think in years past, like I felt that Google was a great, great, great infrastructure company behind the scenes, but never was great at the speed that everybody else was of productizing everything, right? They productize Gmail, that’s great.
and they productize, know, calendaring, that’s great. But there were all these other great things that they had, they just never quite figured out how to deliver them at scale and productize them at scale while Microsoft has done this, AWS has done this, others have done this. And I think that, I think it’s different now. I think it’s dramatically different now, especially infrastructure side. I feel like they’re getting more disciplined on product, right? Like to your point, they have the highest abandon rate. You would see something come out and then like,
oh, that’s really cool. And they’re like, where to go? They’re like, ah, they cut it. Where you look at like Microsoft and now Azure Locals, what, the third or fourth iteration of the same thing that they might tweak, change, maybe even rename, but they stick with their products and deliver it to market until essentially it starts to work out in their favor. AWS seems to be more in the middle of that. They’ll cut some products, but they’ll stick with it. So it is…
it is interesting to see the evolution of Google. One of the things that Google has always been known for, and I’d to get your opinion on this, where they’re at with their partnerships. Because this is where, like as an advisor, this is an important question, right? Like when we say Google and their partners, that means the partnerships to the suppliers that we rely on day in day out to deliver services to our client bases and how we go to market. So where are you seeing that go in?
you know, out of the show, is it as relevant to Google as it has been in the past? Or that was one of the premier things that everybody always talked about. Yeah, love this question because this is just exciting to see. So if I flashback to years past as we’ve watched these hyperscalers build up, you know, their goal in the beginning is to just build great products and they’ll figure out a way to get them to market.
And so early years of AWS at first was not, they didn’t know or understand what a partner community was or is or what that means. now under, know, Matt, Matt Coran, maybe butchered that name, but Matt Coran, he sorely like, he dramatically understands the partner ecosystem and is fully, fully embraced into the partner community. So awesome to see that standard now in the last year or two from AWS. Google under Thomas Curian.
has really, really leaned in and go, oh my gosh, this whole partner community, this is a really big and important thing. And so seeing them embrace the partner ecosystem as a whole, and I know that means a lot of different things, ISVs, SIs, here we are in the TSD side, right, we’ll get recognized as part of that in proximity to the vendors that we talked about earlier. And so I love where they are fully, fully embraced into this partner community.
And you see that, okay, saying it is one thing, but seeing it through the different funding programs that are available, that are accessible to build these out and go reach and get funding potentially to build POCs for our customers, for our opportunities. Excited to see that, you know, we saw that early on in AWS with map funding, know, Microsoft and Azure came in with their version of it. And now to see Google have it, you know, that really tells us that, Google is here to stay.
into the partner community and exciting to see how they will continue to embrace that and continue to put their money where their mouth is, both of trying to get these away from other hyperscalers and see them all compete against each other. that got to follow the money trail. And when the money trail is there, that tells me they’re in. You know, I always refuse to pick a favorite because they all benefits, right? They all do stuff a little bit different. But what.
we’re seeing in particular out of, you just mentioned AWS as well, and Google is really leaning in to the better funding programs that you’ve mentioned and how that trickles in. And then that’s being offered through, you know, in different scopes of works and opportunities and things like that that our advisors are getting into. There’s more to come on this guys. We won’t get too out over our skis, but there is a lot of interest in the modern distribution.
you know, mechanism that you guys are driving is for these, for these hyperscalers as well. And you’re going to see a lot more interest in collaboration through the suppliers and everything that we already, sorry, already have access to. I don’t know where I trailed off on that one. Granted it is Monday guys. Yeah. In case you want to record this a little bit of a struggle bus for both of us. But Josh, if you had one major, you know,
thing from the breakout sessions and everything else you saw, you know, at the next conference, what would that be and why, and like, what do you think would be the most impactful for advisors to pay attention to? Four words. Don’t sleep on Google. Maybe you felt like you could do this in years past. Don’t sleep on it. And, you know, look, I, I recognize that some of the things that we’re talking about are leading edge, but some of it is already productized and usable. And so I think
The speed at which even more becomes productized and usable accelerates over the next six to 12 months. I think the run rate, we have to be just fully aware that the run rate, they made it to 50 billion. That’s an exciting place to be. There is no slowing down on that gas pedal. So we’ll continue to stay with these things like the agent to agent protocols, what new adoption of Gemini has, what POC options are out there, things like that.
We’ll continue to bring those things to you and pay attention to how that works in our ecosystem. But when your customer says they have Google and they have infrastructure or they’re considering Google infrastructure for X, Y, Z reason, don’t shy away from that. We absolutely can help you with that. And that wasn’t the answer in years past. So it’s an exciting time to be here to say, we really don’t care what your customers are already invested in. We can absolutely help. And Google is now a deep part of that conversation.
Yeah, and if you start to break down where they’re winning, it used to be like startup companies that would get on Google to start and then kind of expand into Google Cloud. And now they are getting more migrations or hybrid cloud, multi-cloud type of wins to go along with AWS and Azure. Guys, back to the agent again, think. That’s what’s so exciting about this.
Infrastructure management, hybrid cloud, and all of those things are going to become bigger and bigger conversations because there might be a product that sets in one of these hyperscalers that a customer needs access to. Getting it and then bringing it their environment, working it through, mixing it with the private cloud, doing all these things that have been overly complex, they’re not going to get any easier, but they’re going to get more manageable and they are going to get more attractive and by design.
driving customers to these hyperscalers and they’re going to get a lot more at bats for these conversations. So make sure that you’re having them, give your customers the conversation pieces around agentic AI, the roadmap that they’re on, and then all stacks and plays into it. You start with that AI conversation that Josh talked about that pulls back into an infrastructure conversation, a security conversation, a CX conversation, an EX conversation.
And all of a sudden, you got a roadmap worth of conversations to have that turn into opportunities and commissionable dollars. So Josh, this has been great. I think we’ve all learned some stuff about where Google’s at in the market, where they’re going, and any other things that you want to do to put a bow on this thing. Yeah, just a couple thoughts. One fact check. I fact checked myself. Matt Garman.
is the AWS CEO. Realize this after I said it, Matt Garan is a partner of ours out of the East. So Matt, hey, if you’re listening, just that would give you a shout out today. That’s thing one. Thing two is I would say, listen, do you need to be an expert deeply in all of these? No, absolutely not. That’s why we’re here. That’s why we’re here to help. But what I would say is go look up, you know, if you go Google the Google agent to agent protocol, it’s a few pages. It’s not a crazy read. I want everybody to understand that.
and just have a tiny bit of context around what this means so that when we see a little bit of, know, little iterations that come off of these, play with it, understand it. And if you haven’t used Gemini, done some of these searches, played around with some things that are in there and done some of the deep research, poke around in there, just get familiar, just understand how intuitive it is and, you know, we’ll continue to come back on and talk about new evolutions and the next next we go to, we’ll have lots of good stuff to share. Here, I’ll give a little hit.
for the sales people that are listening, sorry, versus the engineers that are listening. Engineers, do exactly what Josh said. Go read it, scope it and all of that. Sales people, go get the doc, hit the summary button on Jim and I and get the key points of it and turn those into talk tracks. Either way, they’ll be impactful, trust me. One of us did one of the things, the other one did the other, and they were equally impactful in my opinion. So Josh, you’ve been a fantastic guest on your own podcast.
I couldn’t thank you enough for being here and allowing me to pinch hit and hosting. So I’ll turn it back over to you to wrap everything up. Awesome. Great work. Love it. Appreciate it. Love having to prepare just a little bit less. This has been nice. So good stuff. Great event. Look forward to the next one. And, you know, we’ll bring this back to the partners. And so as you’re listening, wherever you’re coming to us, just remember.
Apple, Spotify, these drop every Wednesday. Make sure that you go follow, get those notifications. But as always, we appreciate you listening. Coby Phillips hosting today the Google Next recap. I’m your host, Josh Lopresto, SVP of Sales Engineering at Tilleris, and this is Next Level Biz Tech. Until next time.