Thank you, Sheila, and hello, everyone. I'm so glad to be with you today, share what we have been working on recently, and how this will have a positive impact on our customers. As many of you know, we power experiences across websites, commerce, service, and workplace with one single platform used by all of our customers always up to date with an API first approach. AI powered search, recommendations of personalization are at the are the core ingredients we use to make these experiences perform better. If we look at commerce as we mentioned earlier, our customers need to provide the best relevance, but also the ability to measure it and optimize it for business outcomes. It means optimizing our relevance and AI not only for conversion, but also for revenues per visit and down the road profitability. Those two components have to work in in hand for this great balance between relevance and merchandising and ecommerce. On the relevant side, I wanna share something quite excited that is currently being rolled out at our large ecommerce customers. We call it intent aware ranking, and it is based on our product vectors generated using deep learning on behavioral data. Anonymous visitors will get personalized results after a few clicks. As an example, as you can see on the screen, after a few clicks in the session, if you search, you will get results that will be, that will be related to what you've done. If another anonymous user starts browsing in different session or consults, products in different sessions like this one, looks like more winter related, the same surge for Madhuti will get totally different results. We are measuring this with our some of our large customers right now, and the results are quite the early results are quite good. So we're looking forward sharing those results with you a bit later this year. So moving from relevance to merchandising, which is about the controls that we provide for the merchandisers. We have rolled out a new version of our merchandising up that is coming from our acquisition of Qubit. Some of you may remember we, we acquired a company by the name of Qubit at the end of twenty one that was a leader in merchandising for ecommerce. So we integrated those assets into our commerce offering. So the merchandising hub offers campaign management insights and analytics on those campaigns, the ability to manage personalized content recommendations and badging as you saw in the previous screen. But what is new, from, last last January is the ability to deal with product listing pages directly in the merchandising hub. Later on, we are also going to be able to manage and measure profit optimization directly from the merchandising hub. So let's get inside product listing managers. This is an area where merchandisers manage their category pages, also known as product listings. So in this example here, the merchandiser has the built a beige for men sweatshirts. So please see that there's already, some analytics in there that are quite interesting. So the merchandiser wanna apply a few changes here and see if these metrics will, improve. So with a few clicks, the merchandiser has the ability to do boost and vary, add some additional rules on how the results should be presented. So it will define a here, it will define a boosting rule. So, Nike will be boosted for those results. So you see the results will change automatically. And then there's the ability to pin the results. So the merchandiser can do drag and drop and pin results at the top. So with that, it can, fairly easily see if the results are improving or not. And so when we talk about new controls for the merchandiser, that's a great example. I'd like to switch to, another area of commerce, and that's b to b commerce. This is what distributors, wholesalers, suppliers will use to sell to their, to their retailers typically. And one of the challenge that those distributors have is that they have a large number of products in their catalog, but they have different price books per customer. So, basically, each retailer has its own price book has its own pricing per SKUs. And that becomes a problem to make that available in search and even product listing catalogs because this matrix needs to be, needs to be processed and managed. So typically, what, retailers will do today is they will flatten the structure, and you will have if you decide to attach pricing versus queue in this classic solution, then you will have a very a lot a very large structure that quite frankly doesn't scale most of the time and will have the inability also to change, either pricing and individual pricing or some product description. So that for us is not a that for us is not an option. In this example here, five million products with a hundred thousand customers with individual price. This equate this equals five hundred billion items, which is an index that is either unmanageable or super expensive to manage. So we believe there's a better solution, and that's what we have done. We've been able to add up to a hundred thousand price per skew in our index. And by doing that, we have the ability to provide fast search because the index is smaller. And the other example is a five million document index instead of five hundred billion index. We have the ability to do individual field updates. So instead of taking weeks, it can take a few minutes to index a price in the index. And, we also have the ability because the SKUs are only available in one place at the index. Each every each and every SKU are not duplicated in the index. Then we can apply AI and machine learning in a way that is standard and coherent with the rest of our solution. So we are rolling that out into one, very large distributor at this point. The early results are quite exciting, and we are making this available to our other customers as we speak. Now a lot of partners and customers have shared with us their interest with about the large language models because of the release of chat GPT mainly. And we think it's a great opportunity for Coveo. We think that these new technologies offer the ability to do real generative question answering. Basically, what we see, in our customers are questions like this. Right? My robot is not following the battlefield in the map. How can I solve this issue? I'm planning a fishing trip to Florida in June. What is the optimal equipment for beginner? That's would be more of a commerce question. How to connect my laptop to the TV monitor in the London office? So those are the kind of questions that we expect to see more and more out there and that hopefully can be resolved with a long answer instead of a list of results. So today, Coveo does search, and all of those and this is not going away. So all of those queries that are about discovering content, understanding what's out there, refining a theme or refining a way to ask the question, this is what Coveo does for living today. So in this example here, it's predictive query suggest that will complete the query based on the intent of the user and then provide great search result list with results ordered by relevance with the ability to navigate and slice and dice through the result. And it's coming typically from multiple data sources with a wide variety of content. But then this is what also we need to provide. Right? That says what customers expect. How do I add a new bank account fee? That's a question, and that's a generated generated answer. Typically, that would come from a large language model. So on one side, you have search, and that's Coveo with secure connectors on the content of the enterprise. You've got the index in the middle. You've got relevance on the left side. You've got analytics, administration, integration, UI framework. But then what we're starting to see is systems built using large language models inside the enterprise that requires a new vector database, infrastructure, and the ability to do extraction and embedding and so on. Right? And then it can provide you with, I would say, decent answers from a language perspective, but it's got multiple problems. It's different search boxes to start with, so it may confuse people. It's duplicate content and data infrastructure. It's got separate admins and analytics, and, basically, it deals with different set of facts, so it will provide different answers to the same question, which is a problem, obvious. So here's what we plan to do. We are going to add this capability to our own infrastructure. So Coveo will be able to either respond to a classic search query or to an advanced natural language query with retrieval augmented generation, directly built in in the platform. So we will have the back end with all the administration analytics, depth and breadth of freshness of content and security. And on the result side, it will be unified search box for all queries, generated answers on the most relevant paragraphs of content. So we'll apply all of our relevance to, to the, to the content before sending that to the, before sending that to the, to the LLM. It will have personalization and mechanism to protect against hallucination. So we expect to make that available this summer to select customers under a beta program, and we think that's really, the right way to look at this problem. And finally, on the cloud platform and infrastructure, just wanna highlight a few things that we're about to release. So we're launching our Canadian region in the coming weeks, and totally independent region. So for our Canadian customers, this is great news. And we're also, going to launch early in the next quarter a an active active a multi region active active option, between USCs and US Midwest. It will be early in q two. And with that, we are planning to offer a new five nine option that customers, to protect their workloads with a, with a five nine SLA. This is something that will, likely become available that will become available early next order. So with that, I thank you for your time. Back to you, Sheila.