Welcome, everyone. Thank you so much for taking the time to join us today for a very engaging conversation about how SAP, Coveo, and Deloitte come together for personalized CX. My name is Isabelle Tinoco. I am partnerships lead at Coveo. And today, I'm joined by three gentlemen starting, SG. He is chief expert at SAP. Akhil ChatGPT, VP of product specialist lead at Deloitte, and Simoneau, VP of product at Coveo. Thank you guys for joining, and this is gonna be very conversational. We have some questions that we're gonna go through, and, you're gonna get the the different insights. So with that, let's get started. How do you move from organizing around channels to organizing around what customers actually need? And with this, we'll start with you, SG. Alright. Hi, everyone. So I think one of the important things is the way customers are engaging with platforms and experiences are through conversations. But we are all aware of that, what's happened in the last one year. And now how can we take those conversations and empower it through Agentic AI experiences connected to the entire intelligent experience of the whole enterprise? So for example, let's say we take a industrial manufacturer. The customer searches for you know, I need high pressure pumps for a chemical plant. Think about what the AI agents would have to make sense of behind the scene, the industry context, the product intent, the configuration readiness, and then start having those conversations with compatible products. So you're going ahead and doing some technical documents or, actually guiding the customer through the entire configuration journey. Now exposing this whole engagement experience in context to, let's say, the customer or the engineer or the sales support or even providing those insights to a marketer, That's the new world of engaging with customers in a personalized way. Wonderful. Thank you so much. And, Akhil, what's your take on this one? Yeah. Thanks, Rajiv. Thanks for setting the context. Let me start with the honest observation. So most organizations are built around channels in our experience. Either it could be a web channel, mobile channel, a store. Nowadays, it's Agentic channel. Right? Each one has its own team, own budget, and own definition of the success based on the channels. And it may work for certain organization, but in a majority of of the organization, it creates a disconnected customer experience. The customer is not looking for a channel based operation, but customer is looking for a situation where he needs some help. So where every time when customer try to interact with their channel, it the journey start over again, and it creates a friction to the customer, and it creates a problem for end to end customer experience. Our perspective is look the problem as a situation, look a problem as a end to end customer experience, not by channel, but solve the problem as a customer end to end customer experience. Wonderful. Thank you. And, Simon, we'll we'll close this question with you. Sure thing. And and, definitely, I'll take the the tech route here. You know, we're in that kind of new agent tech era, and, you know, we've been pretty much every time that there's a new set of technology that comes in, we always say this one's gonna break the silos. Right? And and this time, it might be true because instead of being, you know, a a a department with their own set of tools and their their specific channel, the department will have access to agents, and these agents will talk to the agents in the other channel. And this is where the silos will break. Right? So if I start a marketing campaign, I'm part of, you know, the retail team. I will need to tell this agent that this also potentially impact the fulfillment system, the return system. There is an after sales element to this campaign as well, and I will have control over all of this through a set of agents with their own security protocol and and conversations. Thank you so much. Great. Let's go to the next question. And for this one, let's discuss what happens when Discovery stops being a navigation tool and actually starts being the connective layer across the entire experience. So, Akhil, what's your take on this one? Yeah. So discovery or we call it as a search are most visited features in the digital experience, but also most underinvested. Because of the gap, the customer always sees a breakdown in the customer journey. Most of the most of the organization thinks that the search is a tool. You give a keyword, and search is gonna respond to that particular keyword. But now search is no more a tool. It's a digital experience capability by enabling AI features, AI capabilities, understanding the customer interactions, understanding the customer insights, and provide a guidance to the customer in the whole experience journey. So in future, search is not going to be a capability. Discovery is not going to be a capability, but it it's a underlying back backbone for the whole customer experience layer. Great point. And, Ajay, what what do you think? How can you build from the I'm in agreement with what Akhil said. Right? Because think about it. It is one of the most underinvested areas. So I would ask everybody to think of discovery and replace that word discovery and call it curiosity. That's what Generative AI has done across the whole board. So so let's say, for example, wholesale distribution. Think about a customer or a sales engineer or a marketer and use cases around bulk order search. Now you're trying to figure out contract pricing and availability. How can you do intelligent reorder recommendations based on history? How can you do real time fulfillment and delivery selections? In all of this, nobody is going to explicitly tell you this is what I need. There will be a search item that's a bulk pricing bearings of five hundred units. But you're trying to understand and detect the intent, the procurement intent, the urgency intent, the repeat buying behavior. And based on that, you're trying to prioritize contract pricing availability and reorder workflows. So it's a whole new experience of looking at curiosity with discovery. Great. Yeah. And, Simon, let's wrap up this I mean, the keyword here, that SG just said is intent. Right? And and since the very beginning, you know, search has always been the place where people state their intent. You know, everything else is browsing and clicking. Search is the only place where you type. Right? And, and and we are when this entire agent take conversational story started, people try to take people away from the search box and into, you know, a chatbot or shopping assistant or anything like that. But but UX wise, this is the place where you start your journey. So whether you're shopping or you're looking for assistance or anything, that search box become an intent box. Right? It becomes the place where you put your intent, and then the agentic element can do the orchestration and brings you to results, brings you to after sales support, brings you to education, to to start your journey. That's really where the whole thing will happen. Perfect. Yeah. I'm building from intent. Let's move on to the next question. So how do you move from responding to a customer's ask to anticipating actually what they need? And, Akil, what's your take? Yeah. I mean, kind of building on top of, the previous questions. Right? Here, something we see consistently. The organizations already have data and customer insights, but are they effectively using it? The data is there, but it is sitting in silos. As long as the organizations are not gonna effectively use the data and insights, they're not gonna provide a effective customer experience to the customers. For example, let me talk about two examples. One, a customer is on the site and comparing the products multiple times. What it it is showing to the organization is customer is in a decision process but but unable to make a decision. Second, a returning buyer coming to the site and trying to explore. We don't need the customer don't need to start from zero, not from the scratch. The organization already know what customer is trying to do and provide the guidance. In these two examples, getting the insights, understanding the insights, guiding the customers to complete the journey is a bigger portion of this whole customer experience layer. Like, as I said, the organizations have a tools. They are collecting the data, but not only collecting the data, understanding the insight of the customer, and driving the driving and guiding the customers to make the decision, that is a key in every enterprise. And that can happen through by enabling AI tools, AI capabilities, AI search platforms, AI enabled the customer journey. Excellent. Yeah. And, Simon, glad to hear That's that's that's what's happening, and the and the data itself is also very siloed. Right? You you have, you know, your your search analytics, You have your listing analytics. You have your time on page, your bounce rate, and they all serve different purpose. Right? And, again, here, I'm I'm I'm going into that new agentic reality where, for example, I get on the search interface. I have facets. I can click on these facets. They will this will trigger a click event and maybe direct, you know, my my search performance in the future. But I can also have an agent that comes in and say, hey. This is a pretty broad query you've just done. Do you wanna start a conversation? Are you sure, you know, what, what kind of product you're looking for? And try to get these users to have a rich interaction, you know, with with, with the system so that what we learn is more than search and clicks and click stream and and bounce and and these things, but it's actually conversation and and back and forth. Excellent. And, AJ, let's close this question with you. I've been in this AI space for over a decade, and I've seen the evolution from machine learning to heuristic workflows to generative AI to AI assistance, and now we are getting into the world of autonomous agents. And so I like to pivot there with this topic. Right? So so let's take an example, high-tech CRM. How can we have AI agents understand what the channel partners are doing when it comes to searching chipsets? How people are downloading integration documents? People are creating scaling support tickets? How do you actually go about understanding all of that as autonomous agents? Understand that there is a need for expansion of and readiness around it or even predict churn risk or even detect expansions of conversations and use that to guide the channel manager into the next best action. The moment I explain this, you need autonomous agents working with a connected data enterprise system and activating that through various AI use cases, but most importantly, embedded into the business and the enterprise architecture of the enterprise. That's when you start getting the real value. Excellent. Thank you for that. Going to the fourth, question. How does personalization so often feel accurate in one moment and then in invisible in another? So for this one, let's start. Yeah. Simon. Yeah. Thank you. Thank you, Isabelle. And I was about to say, you know, that's it just doubling down on what SG just said. Right? It's it's because of the enterprise architecture pretty much. So you have, you know, personalization rules in place or even very, accurate or or highly technical, you know, personalization AI in place, but if in in silo, right, for which is very dependent from the system you have access to. We we've talked about, you know, unifying data, unifying signal for so many years. It's still not really working. Right? So the idea here is how do you preserve, how do you share data, and how do you use, again, here, the concept of autonomous agent, in order to preserve data and to preserve, the personalization and then use it in multiple system, maybe, again, through an agent that impersonates the users. Excellent. Yeah. And, Akhil, what's your take on this one? Yeah. I mean, this is a most common problem we saw in the enterprises. The personalization works beautifully one place, but the second moment it fails. For example, I'm on a b to b application on a commerce site. When I log in, I can see, like, beautiful banners, recommendations, products related to my preferences. But when I try to transact it, it's gonna fail, but not by not having the inventory. So think that in a typical way to be kind of organizations are a setup, creating a transactional flow takes weeks and months. And suddenly, if user is gonna see there is a disconnect with the inventory or your service or your contracts, the whole experience is gonna break. Now if we are living in a agentic world, this particular personalization disconnect can be easily solved by having connected systems, by having connected data, by identifying these flows to have end to end customer view, customer experience, and solve these problems by enabling those identity flows. So the goal is not about smarter personalization. It's more about a reliable personalization. Excellent. And, let's wrap up this question with you. What are your thought? I think Simon and I can cover it so beautifully. I like to elaborate more on let's look at the world of world of direct to consumer and let's say retail. Right? Now if I look at a retailer, and all the application systems you have, you think about it like inventory systems, loyalty systems, point of sales. The consumer data is scattered across all of there, and each of these engines were trying to do personalization based on the data and the processes locked in that application. But now the journey can begin in Gemini and Google Shopping. Right? And how can we understand that same customer's data across purchase history, which is transactional by nature, make sense of the buying intent which has come from probably the point of sales data's prediction, as well as all the search clicks which they have done on various aspects. Look at the entire wallet share and customer for life from a loyalty perspective and have various AI agents recommending that customer throughout that whole journey from a content creation or merchandising or or promotion engines and so on, and yet have that empathy and human in the loop kind of an experience across that whole value chain. And we see that also, as spoke about earlier, in the world of b to b also. So that's the new world of personalization. I think it's engaging with the consumer and the customer with empathy, with a holistic view of their experiences, and giving them that autonomous experience, but so also blending in the human and the loop when required. Thank you. Well, we are coming to an end. So let's discuss how we can make the shift real. And this is a question that we have to wrap up the session. How do capability, platform design, and transformation thinking actually work together rather than in parallel? From this one, let's start with you, Akhil. Yeah. So let me put it, perspective on this. So most of the digital transformation programs fail not because of the bad technology, not because of the bad strategy. It's because the capability, platform design, and the transformation, they don't come together, and they don't work as an integrated team. Right? They work as a panel work streams. One work stream is enabled with AI, tools and technologies, but the other work stream is not. So when they operate in parallel and when they operate disjointed, the transformation program space. That is one. Second, the successful programs are not through doing the whole transformation at one shot, but doing it in smaller pieces. Identifying the minimal minimum viable product and BP, identifying a certain customer section, customer cycle certain customer segmentation or certain geography, proving that the transformation is working fine, and then scaling it to the next level. So that's typically, we see the successful programs start small. They integrate the teams, and they deliver MVP, and then they scale up on it. And as part of this process, enabling the AI tools, latest technologies, enabling the digital transformation that makes this program successful. Right. Yeah. And SG? So I'll use the analogy which I use a lot, which is baking a cake. So so think about AI project or a use case like baking a cake. So let's say, when you want to bake a cake, you need a recipe, which is a business process. You need ingredients, which is data, and you need equipment, which is the enterprise technology infrastructure and architecture. So if I think of a cake, like, let's say you want an autonomous agents in aftermarket parts and services where engineers searching for it to replacement service parts. You're trying to that's the part of the business process around it. That's the recipe. You want to predict parts, inventory, service procedures, contracts, configure price code, part experiences, search experiences. Those are all the data, which is those are the ingredients. And then you will come up with, let's say, now the equipment. Right? The equipment could be agents, which is discovery agent, a service agent, an orchestration agent. But then in all of this design, you can't keep changing your kitchen every time that you have a new use case, which means you have to work with enterprise systems and and partners and implementation experts to design a North Star architecture, which can embrace the new world of autonomous agents, the new world of end to end experiences. Because you will keep baking cakes, and so you need a kitchen where you will be able to start consuming all of those cakes in a very easy manner and enjoy it. I love that analogy. Thank you for sharing. Yeah. And, Simon, let's wrap up this question with you. Yeah. Absolutely. I don't don't know if I can double down on the kitchen analogy, but I really like it. I just wanna close, I think, by saying, you know, there's there's no choice really. What what's happening now is that we're putting in front of our customers. And and I mean here, you know, visitors on a retail site, the buyer on, you know, a b to b a b to b portal or even, you know, anyone visiting self-service support after sales portal. We're putting in front of them anonymous agents, and there's an expectation that these agents will be able to connect to pretty much everything. You know? People will get on the retail side, and, you're thinking that they might be there, you know, for product discovery, but they're asking about the status of their orders. Right? So you need to open up, these silos. And and if the if you don't do it, then, you know, you pretty much your entire strategy will fall down because nobody lives in silos anymore. All of these systems are now interconnected. Interconnected. The data is accessible, and the the the customer expect to discuss in natural language, with these different agents and have the response regardless of where they are or where they're coming from. So that's just to, you know, to double down on on the importance and the urgency of what Akhil and SG just put forward. Excellent. Well, thank you so much, gentlemen. This was a great session. Short and sweet, but very valuable. And I just wanna leave you, everyone, with our panelists here. If you wanna learn more, please reach out. We're gonna be more than happy to also hear from you. So with this, we'll wrap up today, and have a great day, everyone.