 Hi. I'm Max from Spoke AI. We build tools for product and engineering teams to help them with their communication and workflows across different tools. And instead of just pitching our product, today I also wanted to share a little bit about our journey, learnings from our beta so far, and some of the beliefs around how AI will have to evolve from very B2C focused applications that we all know and use right now in B2B to really create value there. So we now live in a world where software understands English, and you've probably all seen examples where you can now use English and previously had to know how to code or write a complicated formula in Excel. And if software now understands English, does that maybe mean that we can also use it to understand context of what's already happening in our work life across our tools? And at the same time, we also live in a world where we use increasingly many tools, right? So this is simplifying a little bit, but we moved from email to Slack to now many of us using 7 to 10 tools every day, which is great, because those tools are super specialized, but it also means that our communication is dispersed across all of these tools. And we're eventually overwhelmed by information in all of these tools. And from what we see, product and engineering teams are actually hit hardest, because they have rather complex workflows across seven to 10 tools every day. The result is most products launch late. 45% assist that from McKinsey. So yeah, I guess you can trust that. And what that means to companies is that they actually lose up to 33% of profit on a six-month delay. And I've seen products be delayed much more than six months, so that's even worse. And we at Spoke believe that more process, chatbots and complex workflow builders cannot fundamentally solve this problem. And I'll show you a couple of examples around why we think that. And I guess everyone here has interacted with AI through chatbot interface, which is fantastic, because it gives a lot of flexibility, right? You can ask basically anything you want. And it's a very established UX, like we've all interacted with chatbots in the past, so we kind of know what to expect. But at the same time, we have to prompt the model. So depending on how we prompt, we might get a lot of different kinds of outputs. And there's always the question around data access. So people are kind of thinking about, hey, can I actually upload data from my Slack or from my JIRA in a work context? Is that OK? And that leads to a lot of people using AI kind of a little bit secretly right now, like in a browser tab hidden away using chatGPT there. So that's the one side, right? This kind of chatbot interaction model. And then the second one that we think about a lot is workflow builders. And who has set up one of these before here? Show of hands? OK, so this is safe here. It's really cool and powerful when it works, right? And you can basically connect any tool, lots of different ones, with a tool like that. But there is a certain setup cost to it. It's not super easy, especially if we look outside our little startup bubble here. And it usually breaks on edge cases. So if you've built one of these before, you know that it can go wrong super quickly, and then it's annoying to fix. Now, I wanted to share a little bit like how we approach this whole topic and what we think about when building specifically for the workflows and product and engineering teams. So we want to build something that's super easy to set up, unlike a workflow builder, right? So it should work out of the box. We want the output to be hyper-personalized very much like it is in the established chatbot interaction that we have with ChatGPT, for example. We want to be proactive rather than reactive. So going from you searching for information to us giving you the right information at the right time when you need it. And we want to be basically integrated across flexible user interfaces and gradually automate the user experience around AI. So what that means is that you don't have the question-answer model that you have with the Chatbot, right? But you also don't have an agent that just does everything on its own, but you kind of get there in steps. And ultimately, we do that to enable our users to make better decisions within their workflows every day. How does that look in practice? So we started by launching an MVP within Slack. And what that did initially, super simple, was just AI summarization within Slack, giving you context in a very noisy place, but with very low friction. So it just takes two clicks to get started, right? And then you can already get some value out of it super fast, have your first summaries. And this then forms the basis for further actions you want to take based on this context. You can then start launching into your workflows, like, for example, take in an action that Spoke suggests. Last but not least, the MVP helped us by building feedback loops early on, right? So our users could, for example, review and rate the summaries that we generate for them, rather than our team doing it, which we did initially. Now we're moving beyond Slack as the MVP and launching the first part of our main platform. And what we do there is we actually filter and label your communication across different tools. So here's an example from a message in Slack where Spoke detects that it requires an action for you to take, and it instantly gives you a quick overview of what's happening in that channel so that you have the context to take that next action. We then, very much in our Slack MVP, want to give you context through summarization. But now we do it across different tools. So no matter whether a conversation is happening in Slack, Figma, Notion, Jira, wherever it might be, we can give you instant context through summarization and then derive action items from there. And here's an example of how that looks, right? So we derive these action items, suggest them to you, and then make it super easy to take those next steps by launching into your different tools and stop copy-pasting information from A to B and having a mess in different places. So this is how the product looks today. It's obviously evolving a lot, right? But it's becoming this communication hub, command center for product and engineering teams to start with, and then hopefully for the rest of the organization as well in the future. At the moment, we're building this with close to 500 companies, which is super exciting. And would obviously love to chat to any of you if you have questions around this, if you have similar problems in your daily work, or if you have opinions around how AI has to develop to really create value in a B to B context. Take a look at spoke.ai. Our product is in beta, free to use. So thank you for your time.