watch from 19:34

It works. So won't stop. Hi. I'm Ollie from Count, and I wanna show you a completely new way that you can see and explore your business using AI. This is the the account canvas. It's built to allow, to give AI the space and the scale it needs to help me and my team dig into our data. So I'm gonna add an agent into the canvas because I wanna get a handle on a few things going on in my pretend ecommerce business. It's made some suggestions, but I'm actually gonna ask it something quite quite specific. Hey. What's going on in Chicago? Now that's a pretty ambiguous question, so it might take the agent a bit of time to fully dig in, but counts a for collaborative space. So I'm gonna invite in Harriet into the Canvas too so we can work through this problem Together. Hey, Harriet. Hey, hi. The the agent's cracking on over here, but let's go off and ask a few other questions. Anything else you think we should dig Why don't we look at the product segment split? Sounds good. Great. Can you make a single dashboard containing three charts over time, looking at profit, sales, and orders, and with each chart split by segment and grouped by month. Oh, nice. That was pretty quick. So one of the great things about, count is that when you're working with the AI, everything it builds is editable directly by us. This allows us to check what the AI is doing and maybe make an edit ourselves pretty quickly. So for example, what's going on here, Harriet? Something's looking pretty dodgy over here with corporate. Oh, yeah. Okay. So I can ask AI to look into this more, or I could just drag this down. I could filter the segment. So I'm looking at just corporate, and then maybe I wanna look at how this looks for different shit modes. Yeah. There we go. Seems to be second Sunday class is the problem here. But rather than dig into this, why don't we go back and see how the other agent's getting on? What? Still going? It's still I think it's because it's failed. Oh, dear. Oh, man. Okay. That was, like, going so relatively well up until, like, the last bit. It's like, oh, nothing's happened nonetheless. Oh, god. Why is it like every time it's it's different every time. It's a classic thing. It's classic problem, isn't it? Two meter. Do more elsewhere. Do more else what do mean do more elsewhere? Like, just hide the seeds, like, spill the. Like wait for longer. I don't know. We discuss something else away for a bit. It's usually further ahead. It starts it off so well. It's like, oh, really Cracking on before we left it. I think let's try it in a new canvas. I'm sure that sometimes if you keep asking questions in an old canvas, it's more likely to fail. Okay. I'm gonna stop the recording. Hi. I'm Ollie from Count, and I wanna show you a completely new way you can see and explore your business using AI. Here we are in the account canvas. The canvas is built to give AI the scale and space to show me me and my team all we can about our business. So I'm gonna add an agent into the canvas. And I'm gonna get, ask it to give it give me a handle on a few things going on in my pretend ecommerce business. Hey. What's going on in Chicago? Now that's a pretty ambiguous question, so I'm gonna leave the a r the agent a bit of time to go really dig into that. Actually a collaborative canvas. So while it's getting on there, I'm gonna invite Harriet into join me so we can work through this problem Together. Hey, Harriet. Hello. I've got count count the cat AI digging into the Chicago issue. I what we should think about in the meantime. There you go. Why don't we look at the product segments split? Sounds good. Great. Make a single dashboard containing three charts over time, looking at profit, sales, and orders with each chart split by segment and group by month. Cool. Here we go. One of the great things about, count is that when you're working with the AI, every object it's building can be edited by humans. So this allows us to look at the data and play with it ourselves. If you wanna do something ourselves. Damn it. We should stop It's all I can. I I mean, continue. Why? So that we practiced the second half as well. Oh, I see. Love it. Yeah. Okay. So nice. Here we go. One of the great things about count is that one of the great things about count is that all the work the AI does can be edited directly by users. This means it's easy for us to check what the AI has done and maybe edit it myself if we wanted to do something quick. So it's Harriet, me and our AI working together as we dig through the problem. Oh, wow. I thought you were gonna be like, alright. What's going on? No. No. No. Oh, I see. Right, start again. Okay. Let's start this off again. Oh, nice. Here we are. One of the great things about Count is that all the work the AI does can be edited by users. So this has been to check what the AI has done, and may make a quick change ourselves rather than asking the AI to do it. It's basically Harriet, me, and the AI working through a problem together. So what do you think Harriet? I won't be seeing corporate over here. I wonder why that's not doing so well. Oh, yeah. Okay. Let's dig in. So I can copy this below. And I can move segment myself to filters so we can see corporate. And then why don't we just dig into I don't know. Let's look at ship mode. Yeah. There we go. It's dropped off a cliff in second standard class. Cool. So can keep going here, but why don't we head back and see how the other agent's getting on? Wow. It's done quite a lot. So we can click on the agent here and see all the things it's been doing. It's been viewing different parts of the data. When it's failed, it's realized and corrected itself. And as it's gone, it's laying out a story. It's showing us the overview of Chicago, looking to category performance. It seemed at profitability crisis and dug in even further. It's looked at the customer segments, and then it's looking at returns analysis and discounts. So as you can see, it's got a really long way and it can be can go even further if we wanted to. So that's count, a collaborative way, a collaborative AI driven way to explore your business with data. I think that can work. There's a completely different layout for last time. Yes. Completely different. And it's still going. The last bite. Shall we try again? Yeah. Let's get a new canvas then, shall we? And just see what I think it is out of this thing every time. It's good to it's it's good. It seem I mean, even doing this, like, relatively painful like this is really interesting, isn't it? Just to, like, use it even consistently and see this change. It's quite it's something about that. It's also useful, I think. So here's another link. Oh, no. It's okay. I found it. Yeah. Even when I'm doing Hi. I'm Ollie from Count, and I wanna show you a completely new way that you can see and explore your business with AI. This is the count canvas. The canvas is built to give AI the space and scale it needs to really help me and my team understand what's in our data. So I've got a pretend ecommerce business, and I wanna ask a few questions. So I'm gonna add an agent into the canvas. I'm gonna ask it a question about Chicago. Hey. What's going on in Chicago? So the agent's now going off and thinking about Chicago. It's a pretty ambiguous question, so it's gonna take the agent a bit of time to really dig in. But because CAM is a collaborative Canvas, I'm gonna invite Harriet into the Canvas too so we can keep working through the problem together. Hey, Harriet. Hey. How are you doing? So I've been asking our agent to go off and look into the Chicago problem, Harriet. Else could we be looking into? Why don't we look at the product segment split? Sounds good. Okay, here we go. Make a single dashboard containing three charts over time looking at profit, sales, and orders with each chart split by segment and grouped by month. Here we go. So one of the really great things about Count is that every object the AI is building is editable by users. So this allows us to go and check exactly what the AI has done and edit it for myself. It is something we wanna move and change quickly. So it's Harriet, me, and Count dot ai working through the problem together. So, Harriet, what do you think we should look into? I think over here in Chicago, corporate is looking pretty suspect here. I wonder what's going on. Right. Okay. Well, let's look what it all the AI's done. Looks like a pretty simple query. I'll copy it below. I'll move segment to the filters so we can just look at that corporate segment And then I'll look into ship mode to see what's going on with shipping mode. Right. There we go. You can see that standard class has really dropped off a cliff. So that seems to be the cause. So we could keep going here if we wanted to. The Harry, AI, and I work through the problem. But instead, let's go back and see what the other the other first agent's done now. Oh, wow. So you can click on the agent and understand all it's been thinking about. It's been writing queries, where it's failed, it's corrected itself. It's been viewing the data and then digging in further and further. And it looks like it's found some critical profitability analysis issues in Chicago. It's got an overview of sales, and then it's broken it down into different levers of the problem. So we can see the full story being painted here. So we could go further than this, but hopefully that gives you a image of how Count is this wonderful collaborative AI driven way to explore your business with data. Hi. I'm Ollie from Cowen, and I wanna show you a completely new way you can use AI to explore and see your business. So this is the count canvas. The canvas is built to give AI the space and the scale it needs to help me and my team explore our business data. So I'm gonna okay. Let me just get this right. So I want to get a handle on a few things going on in my tend ecommerce business. So I'm going to add in an agent. Can add an n. And an agent into the Canvas. And ask it a few ask it to help. Okay. That makes sense. Alright. Should we start? I'll start again. We can always cut it later. Hi. I'm Ollie from Count, and I wanna show you a completely new way to see and explore your business using AI. This is the count canvas. The canvas is being built to give AI the scale and space it needs to help me and my team dig into our data. So I wanna ask a few questions about our pretend ecommerce business. So I'm gonna add an agent into the canvas and ask it a few questions. Hey. I've heard some things about Chicago. Can you tell me what's going on there? Hey. So that's a pretty ambiguous question. So I'm gonna leave the agent a bit of time to really dig in to the numbers. But because count is a collaborative environment, I'm gonna add in Harriet to help me dig into the problem more. Hey, Harriet. Really? I've got our agent over here looking at Chicago, but only what else we could dig into in the meantime. Nice. Why don't we look at the product segment split? Sounds good. Great. Make a single dashboard containing three charts at a time, looking at profit, sales, and orders with each chart split by segment and group by month. Nice. Here we are. So one of the really good things about count is that every object the AI is building is editable by users as well. So we can interrogate what it's done, make sure it's doing the right things, and maybe for small changes, make them ourselves. So what do you think Harriet? I think there's definitely something dodgy going on over here in Chicago with the corporate segment. Yeah, I agree. I could ask the AI to dig into this, but I'm gonna go away and actually edit this myself. I'm gonna drag this chart over here. I'm gonna move the segment to filters so we can just look at corporate. And then instead, I'm gonna add in ship mode so we can see if ship mode's the cause. Yep. You can see that standard class seems to be the big problem here. Yeah. So as you can see, really what we've got here is Harriet and I with our AI working through the problem together. We could keep going, but why don't we go check out what the other agent has done so far? Alright. It's digging in. So we can click on the AI here and see all that it's been doing. It's been doing various things. It's obviously had a bit of a failure mode here, is fixing itself. But it's broken down the problem, which gives a high level overview, and then every part of the issue is giving us each section of the problem solve is doing. And it's gonna keep doing that. But hopefully, just already, you can get a sense of how Count is a collaborative AI driven way to explore your business with data. Hi. I'm Ollie from and I wanna show you a completely new way you can see and explore your business using AI. So this is the Count Canvas. The canvas is built to give AI the space and scale it needs to really help me and my team dig into our data. So I want to get a handle on a few things in my pretend ecommerce business. So I'm gonna add an agent into the canvas and ask it a few questions. Hey. I've heard Chicago isn't doing so well. What's going on? And that's a pretty ambiguous question, so we're gonna give the agent a bit of time to start digging in and finding out what's going on. But because Counter Collaborative Canvas, I'm gonna invite Harriet in so we can start solving the problem a different way. Hey Harriet. Hey, Louie. So I've got the agent here working away. It's already bringing back some results, but what else can we be looking at in the meantime? Nice. Why don't we look at the product segment split? Sounds good. Great. Make a single dashboard containing three charts over time, looking at profit, sales, and orders with each chart split by segment and group by month. Oh, nice. Here we are. So one of the really great things about Count is that every object the AI is building is editable by users as well. That means you can come into the canvas and click on what the AI is doing, See the query it's written, make sure it's right, and maybe make some small tweaks yourself. So, Harriet, what should we look into more? I think there's something pretty suspicious going on over here in Chicago in the corporate segment. Right. So I could get the I AI to dig into this, but why don't we just drag this chart up here and do a bit of exploration ourselves? So I can see here, I'll move segment down to filters, and then maybe I'll split this by ship mode instead and see if that's the cause. Yep. Looks like standard class is the issue here. So pretty suspicious. Now we could keep digging into this particular problem, but why don't we go back to the original agent and see how it's doing? Okay. So it's it's still going here. We can see it spinning. And we can look exactly what's going on if we click on the agent. We can see it's edited the canvas, added some stuff. It's actually had an error, so it's gone away and fixed the problem, and now it's building more charts. So we can see all its audits a bit all its all its trail here. In fact, it's editable by anyone in the in the in the canvas. And it's still going along here. So we can keep going with this, but hopefully this gives you an idea of how Count is a collaborative AI driven way explore your business data. It did not do very well that time with the thing it returned. Oh, dear me. I mean, that was a very close. That was very close. I reckon everything else about it. You spoke the content I thought was good that time, just if this had loaded a bit more. It's just not even trying. Come on. Clearly, I can't say I've heard Chicago isn't doing well, what's going on? Because that clearly is not. Not the answer. Maybe that's too conversational for it. Maybe it does need to be. So it's like what's going on with Chicago? What do I originally say? Why is Chicago doing badly? What's going on with Chicago is what I asked. Okay. What's going on with Chicago? I think. Yeah. Okay. Well, let's try again straight away because I think it was all fine otherwise. Thank you. It's really good to have someone else. Normally I'm doing this at like midnight by myself and just going absolutely nowhere. But it's a whole day, isn't it? It's not small. Yeah. I don't really use such a nightmare. Okay cool, and I think we should both remember as well to be happy again, because maybe this will be the last one. Maybe. Hi. I'm Ollie from Count, and I wanna show you a completely new way that you can see and explore your business using AI. So this is the count canvas. The canvas is built to give AI the space and the scale it needs to help me and my team dig into our data. So I wanna understand, a few things going on in our pretend ecommerce business, so I'm gonna add an agent into the canvas and ask it a question. Hey. What's going on in Chicago? So that's a pretty ambiguous question. So I'm gonna leave the agent a bit of time to go away and dig into the data and find out what's happening. But because Count's a collaborative canvas, I'm gonna invite Harriet into it so we can keep working through the problem. Hey, Harriet. How are doing? Really. So I've got the agent, Harriet, working through the, the Chicago issue. What else could we be exploring in the meantime? Nice. Why don't we look at the product segment split? Let's do it. Great. Let me come up here, And I will ask, make a single dashboard containing three charts over time, looking at profit, sales, and orders with each chart split by segment, and group by month. Nice. Here we are. So one of the the great things about working with Count is that all the objects the AI builds is editable by users. This means you can click on any of the object the AI's built, see exactly what it's done to get to the answer, and maybe make a small few small tweaks yourselves. So Harry, what should we look into next? I think there's something pretty suspicious going on over here in Chicago with the corporate segment. Yeah. I'd agree. Okay. So I could ask the AI to dig into this more, or I can just simply copy the chart out, move segment down into filters. And then maybe just see if ship mode might be a cause. Yep. Looks like something about standard class isn't quite right. So as you can see, it's Harriet and me and our AI working through the problem together. But we could keep going here, but why don't we go back and see what the first agent is up to? Wow. It's got a lot of stuff going on here. So we can click on the agent and see all it's been doing. It's been viewing different results. It's been asking different questions. And as it's been going, it's been laying out its evidence as we've gone. So it's looked at technology driving profits, and it's obviously got a bit deeper and broken it down to look at returns and product performance. And it's also looking at customer segments as well to see which ones aren't profitable. And it's gonna keep going through all that. But hopefully, can get a sense of how count is a collaborative AI driven way to explore your business with data.