Google Cloud Next 2026: Why AI Orchestration Changes Everything

[00:00:04] Welcome back to the Tech Talks Daily podcast where this episode is recorded at Google Cloud Next in Las Vegas, where one phrase seems to be following me everywhere I go. Agentic AI. Yep, every keynote, every hallway conversation, every executive briefing seems to come back to the same question. How do businesses move beyond AI pilots, actually turn this technology into something that is delivering measurable value?

[00:00:33] Because let's be honest, a lot of organisations spent the last year experimenting with chatbots, co-pilots, expensive proofs of concepts that looked impressive but did very little for the bottom line. But this year it feels very differently because the conversation has shifted from novelty to measurable business outcomes.

[00:00:54] From productivity promises to revenue impact and from isolated tools to fully orchestrated systems that can actually reshape how businesses operate. And today's guest sits right at the centre of this story. His name's Granville Valentine and he's the managing director of AI Go to Market for North America at Google Cloud.

[00:01:17] And he leads Google's organisation of AI specialists and engineering teams that all work directly with executives across North America on adopting Google's latest AI capabilities. I mean everything from Gemini Enterprise to customer experience platforms and applied AI services. Because his team works with some of Google Cloud's largest and earliest adopters of generative AI.

[00:01:44] And helping them move from experimentation to production and most importantly at scale. So today we'll talk about why orchestration matters more than standalone chatbots. How companies like the Home Depot are using multi-agent systems to create real customer outcomes. And why data strategy is still one of the biggest blockers to enterprise AI adoption.

[00:02:10] And what business leaders should do if they feel overwhelmed by the pace of change. As he will explain today the goal is not simply to add AI. It's to rethink the entire customer journey around outcomes. So we'll get into one of the biggest shifts happening right now. The move from reactive customer service to proactive digital concierges. Where search, commerce, service and personalisation all start working as one seamless experience. Rather than a collection of disconnected apps.

[00:02:40] So this is a conversation that will cut through the hype. Get into the practical reality of what enterprise AI actually looks like right now. As someone that records 65 plus interviews a month. I've personally seen a huge increase in browser based attacks over the past year. Whether that be phishing, malicious extensions, account takeovers. The list is long. And it's all happening where people spend most of their time. Inside the browser.

[00:03:08] So NordLayer's new business browser. That's built to address exactly that. It blocks malicious sites before they load. It limits risky behaviours like uncontrolled downloads or data sharing. And gives you visibility into how your team interacts with web apps. And it also helps you stay compliant by controlling access and enforcing policies. Without the need to rely on multiple disconnected tools.

[00:03:33] So for anyone listening that is thinking seriously about reducing risk in SaaS heavy environments. This feels like a smarter and more focused approach. And you can learn more about it by visiting NordLayer.com slash browser. Let me know what you think. But now buckle up and hold on tight. Because I'm going to beam your ears all the way to the show floor here. Where you can join myself and Granville in conversation now.

[00:04:00] So a massive thank you for joining me here. Can you tell everyone listening a little about who you are and what you do? Sure. Yeah, I'm Granville Valentine. I look after our AI go to market team. So we started in 2024. We launched a standalone global function of AI specialists, customer engineers and forward deployed engineers. Who focus just specifically on Google cloud driven AI.

[00:04:26] So Gemini and the surrounding services that includes all of Gemini Enterprise, Gemini Enterprise for CX. As well as our applied AI offerings. And ultimately we work with clients all day. So we're helping from the process of evangelizing how you can use Gen AI through to scaling to production. And we've all been on quite a journey. I mean, only last year, I think many companies found that simple chat bots weren't actually moving the needle on the bottom line. As a lot of expensive AI projects as well.

[00:04:56] But how does moving to an orchestration based model help businesses finally turn AI into a genuine profit driver rather than just an experiment? And the reason I asked that we've heard so many stories about the ROI and pilot purgatory. So tell me what you're saying here. Yeah. I mean, it comes back to the old outage of is it technology looking for a solution or is it an outcome that we're working back to use technology to solve? Right. And multi-agent orchestration is about starting with an outcome.

[00:05:23] It's whether we're growing user adoption, whether we're providing a personal shopper type of experience that's multi-threaded and delivers delight, but also upsells the customer and provides them kind of the right products to purchase and expand their shopping cart with. That outcome is best driven by multiple agents working together. And examples of that can be agents.

[00:05:48] Maybe one is thinking about the user experience all the way from start to finish. Maybe one is thinking about a very specific type of product and how to position it. Maybe one is thinking about upsell. Maybe one is thinking about cross-sell. Maybe one is thinking about checkout. And one is customer satisfaction. But by having all those different agents kind of working together to provide a really unified journey flow, we just ultimately get better outcomes.

[00:06:17] Orchestration is also about leveraging the right data to power that journey. So having a context that is used to bring the right data in at the right times along that journey flow is obviously going to be really important. If we go back to those chatbots of old, right? You just had a simple LLM providing a multi-turn experience with a single direct connect through RAG, right? That was the data connection.

[00:06:45] That's ultimately not a very polished journey because you're just going to get out whatever happens to come from the data powered by an LLM turn. But that full multi-agent system is going to do the full kind of job delivering the most polished outcome for the customer. And it feels like there's been a big shift in attitudes this year. Maybe it's the agentic thing, but it's no longer just about increasing productivity. It's about real business outcomes and driving value. Are you seeing that shift as well right across the board? Yeah, absolutely.

[00:07:15] So let's look at just an example just to kind of give you, to bring it to life. Home Depot with us worked with us on the Magic Apron app. So if you go to homedepot.com right now, you know, it's a similar experience. If you walked into a Home Depot store, you love having that orange apron, the individual who's going to meet you at the door and help you with the project, right? If I'm going in to build a birdhouse with my kids, ultimately I have a premise, but I need a lot of help.

[00:07:45] What is the right wood? What are the right tools? And I'm going to have a lot of questions along the way through that journey. So the Magic Apron agent for us is a multi-agent system built in Gemini Enterprise for CX. And it provides that full personal shopper journey that then delivers delight and improved net promoter score, which is quantifiable from the end user.

[00:08:06] But certainly more, you know, more upsell, more fully delivered customers through the shopping cart and increased ROI. Yeah, I think we're also starting to see a shift from reactive customer service to proactive digital concierges. A few days ago, I was talking with Virgin Atlantic.

[00:08:29] They'd created one and yes, it was solving a number of problems, but the big hit was allowing this concierge was allowing them to map their own journeys and create different trips, etc. So how does unifying search commerce and service into almost a single autonomous journey? How's this fundamentally changing the way a person interacts with a brand compared to the maybe fragmented apps that we use today?

[00:08:53] I think the advent of consumer tools is shaping consumer behavior. It's shaping what we expect of what quote unquote search would have provided years ago. And that means the consumer is willing to go on a nonlinear journey to just find a link.

[00:09:16] So what maybe when a consumer starts their journey through what would typically be a search bar, it can start to take a different shape along the path. And it can start with just asking a simple question, but it could be planning an entire vacation. It could be planning the adventures that one is going to take during that vacation. And weather might be important. Geolocation is going to be important.

[00:09:44] And the images of maybe those adventures or the menu items of the meals and the restaurants along the way is certainly going to be very important. And that's a combination of search, of a personalized journey, of grounding with Google Maps, of grounding with Google Images, of Gemini Enterprise for CX shopper flows, all delivering one unified process and experience. And I suspect there will be business leaders listening that feel somewhat overwhelmed by this.

[00:10:14] They've relied on a website and traditional search, but now they have to think bigger and to create what we're talking about here. Any advice you'd give to those people listening on where they should be starting? Number one, it's very helpful to have a core innovation partner who can educate you not only on what's available today, but what's coming.

[00:10:34] And so we try to play that role, obviously, and spend a lot of time with clients thinking through the reshaping of the way that they can digitally meet their customers where they are. And there's a lot of experimentation. So being open to trying one thing and knowing that it will change in the future is a super important new advent. This is not an application you're going to deliver, and it's just going to be finished and complete for your end user.

[00:11:03] It's going to be a process that's going to evolve as new modalities, as new agentic capabilities, and new models become more available and higher fidelity to that process and delivering that outcome. So ultimately, my advice is get a great innovator in your court, someone who you can trust and have an ongoing dialogue, deep partnership with, and then start experimenting and start learning.

[00:11:29] One of the first things that we see on the journey of being able to go into full multi-agent systems, it starts with fluency. Fluency to me means you just know how to use the tools, even for your daily life, to provide value and to learn from them, use them as thought partners, and to take action.

[00:11:51] If you can start by using the tools yourself and then build into building your own agents for your own personal use, and then do so in a professional setting, sharing agents with friends and coworkers and driving kind of tactical outcomes, then you will graduate into more advanced skills to be able to drive these multi-agent systems that then you have the confidence to externalize and bring your own customers on that journey with you.

[00:14:19] A quick thank you to Danodo for partnering with the Tech Talks Network, helping us deliver more than 60 expert interviews every month. And if your teams are tired of waiting on complex pipelines, this message is for you. Yeah, it's time to stop waiting on those complex pipelines, because Danodo can help you publish high-quality data products faster.

[00:14:44] And do so with built-in governance, semantic consistency, and self-service access. So if you're looking to learn more about how you can turn data into impact, check out Danodo. And they're easy to find. Simply visit Danodo.com. Although the big buzz this year is around agentic AI, there's still a lot of buzz around generative video and photo editing, especially for marketing.

[00:15:10] But beyond just flashy content, how do these media tools actually, or how are they being used to drive core business metrics like customer retention or order volume? Because there's that big focus on ROI this year, isn't there? Absolutely. I think if you looked at something as simple as food ordering, bringing a still image of a 2D piece of food into a three-dimensional with steam coming off, with a human interacting with it.

[00:15:39] This is all creating a much higher level of engagement. Home shopping, being able to rethink the walkthrough. Agents spend a lot of money even just editing the walkthroughs of homes and marketing that content. And ultimately, they're responsible for suddenly becoming marketers. So making that job easier, making the content itself higher quality, higher fidelity, better editing capabilities.

[00:16:05] These are types of things that we can do that not only create a better outcome, but save the broker money from having to go through all that in a manual process. And you take that across e-commerce, you could take that into ad copy creation, into movie development.

[00:16:26] Ultimately, we're providing tools that are helpful extensions to create more with a similar budget, to extend ad copy, to provide special effects that maybe the budget didn't have room for before. And to be able to tell stories in a whole new way, whether that's about a simple piece of food or a full feature film.

[00:16:54] And as someone right in the heart of this space, when you're here at the event, what are the big announcements that excite you this week? There's so much happening here, so much going on, but what excites you? Maybe it's the conversations or all the announcements, but what was it that really excites you about it here? You know, one big one, we talked about the Apple partnership.

[00:17:12] For me, I've been an Apple power user for many years and the distribution of the Gemini model, the engagement that we're going to get through that ecosystem is really exciting. We've seen a great partnership with the Android ecosystem and the Samsung type partnerships in the past and extending that to get Gemini in the hands of end users is really exciting. I was especially excited about Gemini Enterprise and the Agentic capabilities.

[00:17:42] I think that that message landed really well with the executives that are here because they're all on this Agentic journey. They're excited about the productivity app and the capabilities that are native to Gemini Enterprise, but now being able to tackle the full Agentic transformation of their enterprise with a consistent tooling, with the ability to monitor and manage and optimize not just the Google built agents,

[00:18:09] but third party agents and have a whole marketplace to access those agents as well is tremendously exciting for our customers. And then the models themselves are continuing to provide more native capability. There was a model update today around the live API, which also had native avatar capability. So voice in as a modality is super important because of the intonations that it picks up,

[00:18:38] because of the linguistics that are added, even the sentimentality and emotionality that comes through when the model itself is ingesting the raw voice versus going voice to text to the model. And then being able to pair that with a live avatar will really power a whole new paradigm of personalized and human-like experiences in the customer experience suite. And it really is an exciting time. So many opportunities, but the pace of change continues to move at breakneck speed.

[00:19:06] But there's also that concern that, hey, it might never move this slow again. It's going to keep going. So if we were to sit down next year at Google Cloud, next 2027, what do you expect to have changed most dramatically in how enterprises are using AI day-to-day? Any predictions you'd like to make there? If we looked across multimodality, I think the fidelity and controls that you can get over the different mediums is going to increase materially.

[00:19:36] So when we talked about media and the ability to use that in a business setting and create more engagement, the editing capabilities, the ability to craft the message and the story natively inside of the tool is going to see a step function improvement. The other piece is just the fidelity around agentic action-taking in the enterprise will have a massive step function over the course of the next year.

[00:20:04] So being able to take my intent and expect the agent to deliver all the way through to the output that I'm asking it to achieve while it orchestrates on my behalf, while it works with other agents on my behalf, that in the enterprise is going to be game-changing and will reshape how enterprises structure themselves, what they expect of their human workers

[00:20:29] and what they expect of their agent workers inside of the organization and of their partner organizations as well. So ultimately, I think that combination of agentic capability that will be stepped up in the native to the models, as well as the core controls inside of the modalities of video and image are just going to be massive breakthroughs.

[00:20:55] And it's hard to even imagine all of the breakthroughs that we're going to see on the customer side. I think that is a thought-provoking moment to end on. And for everyone listening, I'll include links to all the announcements and indeed your LinkedIn if anyone would like to connect. But I know how busy you are, so thank you for speaking with me today. Happy to be here. Thanks for having me. One of the things I loved about this conversation is that he brought the discussion back to something many business leaders desperately need right now. Clarity.

[00:21:22] Because yes, there is no shortage of AI ambition in 2026, but there is still a huge gap between ambition and execution. Too many companies are stuck in what people politely call pilot mode and what most CFOs probably call expensive frustration. And his point about starting with outcomes rather than technology is probably the most important lesson of all. Businesses don't need just another chatbot.

[00:21:51] They need measurable improvements in customer experience, revenue growth, operational efficiency and trust. And this is where orchestration changes the game. Because multiple agents working together around a real business objective, this is something that creates something far more valuable than an isolated tool. And I also thought his Home Depot example was powerful too, because it showed how AI becomes useful when it feels natural.

[00:22:21] Because customers, they're not asking for agentic systems. They simply want someone or something that helps them solve a problem quickly and confidently. And that can be where AI stops being a tech project and starts becoming more of a business advantage. And his advice for leaders there, feeling overwhelmed, was exactly right too. Start using the tools for yourself. Fluency matters. And the companies that win here will not be the ones with the biggest AI budgets.

[00:22:50] It will be the ones willing to experiment, learn fast, build confidence over time. And maybe the biggest takeaway from the conversation is this. AI is changing from a productivity tool into a structural business decision. And it will shape how companies organize teams, serve customers and define competitive advantage. So a big thank you to Granville for joining me here at Google Cloud Next

[00:23:16] and for helping me and hopefully you help make sense of one of the biggest conversations that is happening in enterprise technology right now. But over to you. Where is your business on this journey that we talked about today? Are you still experimenting with AI? Or are you already building the agentic enterprise? Let me know your thoughts. Hop over to techtalksnetwork.com. Lots of ways of getting in contact there. And I will speak with you all again tomorrow. Bye for now.

[00:23:48] Bye for now.