Dieser Inhalt ist nur in englischer Sprache verfügbar.

Well, good morning, everyone. And I want to start by thanking our partners, Proficient, and italent two great partners for probably more than a decade working with us at Coveo. But really all of you in the room, the vast majority of you are customers of Coveo. And this is a networking event, so you know who's in the room. This is about sharing. We want to show you certain things. Obviously, but it's about networking, sharing and sharing your own experiences and so on. And we at Coveo want to be a catalyst for that. So welcome to all of you. So today, this is my agenda. Essentially, I'm gonna go through a little bit about Coveo, but not that much, but really talk about the fact that how we position, how we see the evolution of search, chat, conversation, and how we all make it work together. And really our view for gluing it all together what we call the intelligent connected customer experience or user or employee experience. Very briefly about Coveo. This is kind of the status where we are. We are now five quarters into being a Canadian public company. I will say proudly that we beat down list expectations every single quarter. Since our IPO. We're also a pledge one percent company, by the way, where we pledge not only one percent of our time and and we pledge one percent of the equity of the company when we went public, and one percent of our technology, actually, which in this case is deployed to the cause of democratizing access to knowledge and education across the world. We think six billion people on this planet don't have or are not born with equal opportunity to knowledge and education, and that's our cause, and we're very committed to that. You guys all know what we're doing. And the big news today is part of the Coveo platform we're adding generative AI. We actually believe that it all has. To work together. And we're going to demonstrate why and explain why today, so more to come. And obviously, Coveo is a very proud Salesforce pretty important, I believe. Salesforce people in the room. Thank you for being here. We are a so called submit partner of Sales force well, so an SAP endorsed partner worldwide and Adobe Gold Partner and recently gained the ISO twenty seven thousand one certification, which is on top of our SOC two type two and HIPAA compliance, very important to us. And of course, we benefit from some industry recognition, but The net of it is we have eight hundred people in the company, and all we do, all we've been doing since two thousand eleven is deploy machine learning in digital experiences, and that's we have one single platform that does that across three lines of business, which are customer service, commerce, and platform. So, we believe that we are recognized industry leaders in applied AI. We don't have Foundational AI. We were not an LLM and etcetera, but we've been using LLMs at Coveo for three years now. Some of you are using some of our technology named smart snippet case, classification, and all that. So we're quite familiar with that, and we work exclusively with leading global brands. That's why we exist. We exist to accelerate the application of AI platforms, and in a way that helps our algorithms, our I'm not only to personalize, but to profitize. You won't find that in the dictionary, profitize every experience. So, it's about maintaining that balance, between the personalization, how you drive more relevance and personalization, but doing it in such a way that actually reduces costs, accelerates revenue, and so on. So for more than a decade, we've been doing that, and that's what we call relevance at scale with our platform. And for us, when we see all that I kind of hate, by the way, the term hyping away, but there's a lot of noise. Let's put it that way. And for a company Coveo, you know, one of the things I said to the employees recently is it's hard to play strategic areas in the middle of a YouTube concert. So our strategy has been last to hype first to result. In fact. We've been helping companies with delivering recommendations in session, ranking, in session, dynamic navigation, but more importantly, securely from ubiquitous data sources and read that last sentence here from multiple data sources, that's going to be critically important. In the world of generative AI, the ability to access multiple sources compounds the value of generative AI. And we're going to talk about that. So our goal is to deliver these connected experiences that, as I said, maximize business outcomes. So we understand the problem. And we've made three important announcements last quarter. Number one is we've made obviously what we're going to talk at length today. We've made the GenAI, our Gen AI, which we call for a reason, relevance generative answering as as a capability because it is about relevance. We've made the announcement of our SAP global partnership in Commerce, which is which is extremely important. We've been working with SAP for a long time, and this is a global partnership that we've deployed with them. And we've made a ton of announcements around cutting edge innovations a few years ago. We bought a company named Cubit in the UK. That does AI powered merchandising for customers like Louis Vuitton Gucci, stay louder, Kate's paid, mac cosmetics, etcetera. Which is really in session, dynamic optimization of commerce experiences to maximize profits, and really empowering. And we released the end street first merchandising hub for to empower merchandisers to perform commerce based on data and outcomes and optimization not instinct. Of course, we've announced a number of other features. So that's basically quickly quick overview of where we are and what we do and what we've been doing. Look, no news to this audience, personalization, data and AI are changing shaping what we call the experience economy. Forget digitalistic table stakes. You had to think about digital a decade ago. This is the experience economy and it's powered by AI. And their very reason, you know all that is that these competitors have raised the bar. And they've essentially changed entirely the way that people expect, what people expect from digital experiences very individualized experiences and set new levels for these experiences, including now GenAI. GenAI adds this new dimension of conversation, of answers, not just links and and etcetera, and needs to be integrated into that. But let's not forget that these companies also use algorithms that maximize profitability. That again, you know, if you go to Louis Vuitton, And after four clicks, fully anonymous Coveo actually knows that you're going to buy a ten thousand dollars purse. And we can actually change dynamically, the experience accordingly. That's how powerful this is. So fundamentally, people expect experience that are designed just for them, that just for me, oops, sorry, that just for me interaction and that are there are people centric, not contact centric, you know, when Laurent founded Coveo, he said something, and that was back in the days, but in two thousand and eight, two thousand and nine, he said something to me, actually, because I was actually the angel investor originally, not but not, so I didn't create the company. But what's fascinated with Laurent. One of the things he said to me in two thousand eight, two thousand nine is he said, watch this company Netflix. Watch this company, Amazon. Watch this company, you know, they are people centric. They use this thing named machine learning because machine learning is not new, by the way. It goes back, you know, more than a decade. And of course, it wasn't as advanced as today. But it's about being people centric, not content centric. Search, which was the origin of Coveo. Search traditionally was content centric. Flip it on its head one hundred and eighty degrees, understand the person at the other end. You've become people centric. That's what Coveo did, essentially, and started using this thing named learning, I will tell you that for a decade, we were, you know, actually in the early days, we were a little bit preaching in the desert, because people didn't quite understand the difference between search and AI, and people would say, Coveo was a search company. We would say, well, yeah, but not quite. And so the ability to do this, now do it in conversational way you'll see it today. Create experiences that are prescriptive that understand intent and that our coherent across the entire journey is really what people expect. And the good news for us and for this industry is that enterprises, now mainstream enterprises, of course, many of you are tech companies. So by design, you understand tech better than most. But mainstream companies now understand that AI is behind these types of experiences, and that AI is absolutely a game changer. What our customers have in common, whether you're Pfizer, you're Thompson Reuters, you're a large retailer, or you're, you know, a tech company, more more of you in the room. What they have what they all have in common is they're all large enterprises, and they deal with massive amount of content or products, I would say, compounded with a very large and diversified audience. And the only AI can solve this problem at scale. You can't make money. You can deliver highly individualized content or product experiences in commerce or content experiences in search and service, you know, with at scale, to this kind of volume of audience unless you use AI. Bar none, it is that binary. Of course, GPT is bringing a radical change changes the expectations. It increases the expectation, but and and and to us, it's a make or break. You know, AI for all companies, not just tech, you know, if you make garage door openers, which some of our customers do, or if you're a manufacturer or distributor of truck parts and so on, AI, you know, you either you you either will be a business who adopts AI right now, or you know, you will be one who will compete against someone who does. And I can assure you that it's AI or die. Meaning that you're not gonna win against, you know, AI powered experiences, period. You know, it is that binary, it is that exciting, but it is that that binary at the same time. And the reason is because it brings significant value. In commerce, obviously, you throw AI on a large commerce site, we routinely within ninety days increase sales by ten to fifteen percent when and and we actually impact profitability as well. Routinely. In customer service, many of you use us for customer service, we drive more self-service intelligence, more deflection, lower costs, increase customer engagement and customer adoption, in fact. So if customers are happy self serving, they buy more, has a ramification, and you can connect both. And within our platform business, within search, it's about proficiency, it's about proficiency gains, given the right information to the right employee at the right time. Is much better than the efficiency is doing more of the same. Proficiency is about people doing more on their own, much more important if you're if if you're a business, in fact. And so this is the power of that technology. Because GPT has been such a game an experienced game changer that suddenly you know, no other technology has been that fast to put to to get the paradigm across to a hundred million users, right, than than Chad GPT. It's it's it's a game changer. It's a and that's driven by quantum leaps in large language models. Now open AI is not the only large language model. Obviously. And as we said, we've been using a few over the years, but it's it's really this year really marks the beginning of of a new era. The one thing we want to get to get across today is that, GenAI is not a separate channel. It cannot be treated in a siloed way. By the way, search is not going away. And you cannot have a search channel, and it's a chat channel that will provide a different answer. It needs to be combined in fact. It's the same data, it's the same user, it's the same content. And of course, it might sound self serving, but we'll actually show that to you today, and how that works. In our view, the worlds of intelligence search, recommendations, chats, generative chats, conversation are all converging in a new digital experience paradigm, a new and better. That is conversational, prescriptive, advisory in nature, And and essentially, it changes, you know, what what people expect from these from these these experiences. So for us, it's it's it's been a continuum in fact. If you look at the continuum of search, you know, and we are a search company, we, I would argue, we are the leading advanced search company in the enterprise globally, certainly recognized that way. Certainly, the fastest growing and and so on, and we've been at it for for a long time. The real and this has been the evolution, right, from keyword search to federated which didn't work all that well, to unified search. You know, Coveo was Gartner Topwright in Enterprise Search in two thousand ten. And that that was because of unified contextual relevance, essentially. We understood how to rank a knowledge article from one source versus a YouTube versus a YouTube video versus other sources. And essentially, nobody else understood how to do that. What's the most because we understood the person in the context of the other end. But the game changer for us was really AI. And the fact, you know, when we started working in deep learning, in machine learning, and in an AI semantics. And now, obviously, we think that the future is advisory And again, we're going to show that today. But I would say the most powerful thing about AI is it learns from every interaction. If it's done well, it learns from every interaction to serve the next. And that's really, really powerful. It's just like when you log on Netflix. Netflix asks one the most important question Netflix asked. Who's watching? That's the only question that matters. Who's watching? And then based on AI, Netflix assembles an experience that is designed just for you. It is so good that you don't need to search. You can search Merrill Street, you don't need to. It creates an experience for you based on what it observes. Never third party data, always ethical, and so on. Netflix, I believe, is very ethical company. I hate Facebook by the way with a passion. I don't like the ethics of that company. I'd say it, you know, very clearly because they go behind your back to find third party data, you know, to understand your orientation, your ethnicity, your age, and and etcetera. Nobody should ever do that. But observing a user, there's nothing wrong with that, given that permission, especially if the user can opt out, that's GDPR. And so essentially, that ability, in fact, of a Netflix to say, okay, given what I understand the hundreds of millions of other users, do, given I understand who you are, what you do, what you watched, what you stopped watching, what you dropped, what you browsed. Here is an experience for you, you don't even need to surf. To search. It's a prescriptive AI powered experience. And this has been the journey at Coveo, our journey through machine learning and deep learning and in everything we've released over the years. Now, here's the punch line, you know, on large language models and conversational AI. It is a relevance problem. Big brands can't hallucinate. It doesn't work that way, period. So combining LLMs with mature and reliable relevance capabilities is app is an imperative, and that's our point of view, and it's binary. It's an imperative to deliver on generative AI that enterprises can trust. If you can't do that, you can't get there. So as I said, we've worked on that for a decade. We didn't work on GenAI for a decade. We worked with large language models now for three years. But we've worked on relevance for a decade, in fact. So what we aim to do is modernizing the enterprise relevance experiences we've always been in that business. And now by adding LLM powered generative and conversational AI. And solve the key challenges. There are nine CIO headaches. So, you know, when Hyde settles, The only thing standing is reality. And here's the reality of Euro CIO. Security, you cannot generate an answer from a piece of information that would be unsecurely accessible by the user at the other end. You cannot breach the permission. Period, that's it. If you're Vanguard, if you're Citibank, if you're Boeing, you can't do that. Privacy relative to public jet engines. No enterprise is going to upload its content into a public jet engine. Enterprise proprietary content for the same reason. And then again, multiple content sources in the era of generative AI, you need to access access or multiple sources of content. And despite the power of Salesforce, despite the power of Microsoft, the Oracle of SAP, of Workday, of ServiceNow, etcetera. Our customers have a little bit of everything, and they have content everywhere. And you need to access that content highly securely at scale in real time. And the power of generative AI increases financially with the ability to access more content, not less. So you can't do just you cannot think about Gen AI just in one silo. You need that layer of software. In fact, that does that. Number five, currency of content. I don't think, you know, I was with the CI of HR block with customer of ours six weeks ago in Kansas City. He walked in the meeting and said, how would you like your tax return to be prepared on last year's tax law. How would you like your tax return to be prepared, you know, using a tax that would have been made up by a generic article. It doesn't work. It doesn't work. Factuality, veracity at scale, coherence in search and chat channels, I don't know what the world is thinking, but search is not going away. So do you want to operate a separate search channel and a chat channel. Do you want those two to be fed from a different source? Give a different answer. Of course not. Sources of truth, verifiability, you need to keep the linkage with the source of truth because companies need compliance. They need to verify. If you're Pfizer, if you're a roast pharmaceutical, you answer to doctors. If you're Boeing, you know, if you're Boeing engineer trying to answer a Singapore airlines, you know, mechanics trying to fix it grounded aircraft in Malaysia, are you gonna you need to verify. You need to trace, in fact. And finally, no one's talking about costs. We thought about it. So so what about the cost? No one right now is talking about the cost. Did you know a cost of a chats about a thousand times the cost of a search query, a thousand. Are you prepared to spend that kind of money? No. There are ways around that. And this is so this is essentially what we've designed. You know, instead of getting in the hyphen, instead of why we put our engineers at it, and we basically built on what we do. We're well positioned because the Coville relevance cloud, when you think about what we do, and what you do with with Cavell, we have a secure we have the ability to to reach content at scale. We have the ability to unify that content and enrich it. But now we use AI. And semantic AI to enrich that content, product catalogs, finding attributes, you know, enriching customer service content, knowledge, etcetera. And then we have the obviously, the platform that delivers best of breed personalization. And and and, you know, based on behavioral and contextual, you know, information, etcetera. That's that's essentially the relevance cloud. That platform generates highly relevant content that you can generate your answers from. So all you have to do is do prompt engineering. Is take the output of the Coveo relative AI plat sorry relevance cloud platform, feed it into LLM. LLM by the way of your choice. The LLM that will be trained on your content. So bring your own LLM. BY LLM is the strategy. Bring your own train it on your data, and then that's Coveo relevance, generative answering that you're going to see today, which is coherent with search channels, with navigation, with the personalization and and etcetera, and is integrated in in the platform. So essentially, we think that enterprises I'm actually ahead of time. That's wonderful. We think that enterprises We think this is the future. We think enterprises will need a spinal AI ability that can reach content at scale securely become the hub that actually generates relevance securely for the connected journey. In order to do that, you need three things. The first thing you need is to securely unify and enrich content. The second thing you need is to be able to generate In session, personalized conversations, personalized results, personalized content, in every app, essentially across every digital journey. And that's precisely what our platform is designed to do. Fundamentally, we think about it in three layers. I didn't make this up. This was Satya Nadella five years ago. There's a records layer. There is a, you know, data records layer. There is an apps layer, which is where people engage, whether it's a portal, a CRM, you know, a commerce site or site, it doesn't matter, an intranet, it doesn't matter. This is the app's layer, this is where people systems of engagement, and there's a layer in between that we call intelligence, the relevance intelligence layer. That's fundamentally the platform. So we're very excited at Coveo because that's fundamentally the platform that's been more than a decade in the making. So essentially, the ability to securely unify index indices and enrich that content is critical. God, we know how to do that very, very well. Across. And that's not a sexy thing, that's plumbing, but it's really, really important plumbing. We have more than fifty people in our connector team at Coveo. They do connectors. Just just Salesforce. I see Salesforce in the room here. Salesforce has seven layers of security. You need to be able to you need to understand how to index that, and how to index that, you know, very, very, very well. So reaching that content, whether it's Salesforce, it's confluence, it's it's it's your ERP content, your databases, etcetera. You need to index it. And the secret sauce of Coveo here has been that we index you know, document level and record level security, we index security, we pull security in the index, and nobody else can do can do that. We tried you know, to you know, over the years, we tried to run Kavailan, elastic and etcetera. There were always some shortcomings. So the quality of that index and the scale, and the control you have over it. Of course, the security, the compliance, and and etcetera, we deal with large banks, financial institutions, etcetera. So infosec matters. So, that's really, really critical. Once you pull it within the index, in fact, you need an integration layer to to bring it up. So what Coveo does in the middle, in fact, is it provides the infrastructure that can do that within essentially any digital experience, any app, and collects the signals along the way. We think that spinal ability is a growing industry. We think that's really, really that's really, really critical to cut across every app and every content source in a very, very flexible fashion, and follow the user, you know, through the journey. If you're in customer service and you're connected to commerce, then you know, how about how about, you know, carrying that signal, and of course, collecting the signal every time just like Netflix does. Every time, you know, what did you watch? Whether it keeps adding. That's the beauty, the beauty of AI. And that's really what the Covayo platform is all about. So when you close the hood at the end of the day, we want to be the platform that powers the intelligent connected experience across the enterprise, and that means essentially being able to serve in real time in session every single interaction and then being able to learn from every single interaction and outcome to carry it forward to the next one. So we are at work, you know, within multiple applications here. You know, whoops, whether it's mobile. Web sites B2B, B2C commerce in product is going to be a major push for us. We think major revolution in customer service is actually going to be about capturing the customer's signals right within the product. Why have customers leave the app? In fact, or the product to have to go to a self-service site. You know, why do we need to do that? If we collect data about usage and etcetera, we can actually automatically be prescriptive and so on within the product. Contact centers, you know, basically the whole service journey and of course, Internet's employee experiences and so on.
Juni 2023

The AI and GenAI Imperative For Enterprise Success and Profitability: Going Beyond the Hype

Beyond the Hype: Build your AI Roadmap
Juni 2023
2023 is clearly the year of AI and a new digital era: an industry-wide transformation that will make or break countless enterprises, those who adopt AI the right way, and those whose competitors will do it first.
The worlds of intelligent search, self-service, conversational and predictive experiences are converging into a new, more modern paradigm across digital journeys, from commerce to service and websites, and in the workplace. Investing in the application of AI is the only way to deliver experiences people expect while competing profitably. Those experiences must be secure, trusted, and your enterprise brand can’t hallucinate.
Join Louis Tetu, CEO and Chairman of Coveo as he discusses the new paradigm we’re entering with GenAI, how AI will optimize experiences and business outcomes in ways not humanly possible, and how your customers, partners and employees expect to engage with you
Louis Têtu
Vorstandsvorsitzender, Coveo