The 5-Minute Demo That Sparked monday.com's AI Revolution | Amichay Even Chen

Hi everyone, I'm Amichai from today from monday.com. I'm a group product manager. You'll see what I do. I'm building basically AI products for enterprise. I worked at Microsoft. I used to work at Axis Security, acquired by HP for half a billion dollars. Recommend if you can work at a place like that, and I'm a European debating champion. This is basically how I met Nick. I beat him. Also, be careful in the Q&A. I want to take you back a little bit. This is even takes even a couple, even more than a year back. Okay, AI just started. Okay, we you open every website and they tell you look. The world is about to change. You're going to lose your job. Duolingo is firing all the translators. We don't need them anymore. SaaS is dead. AI is going to eat us all. It was even before the actual SaaSpocalypse. And I looked at that and I said, okay, another apocalypse. Let's see who— every month I get something like that. And then Like time passes and I'm trying to— and things changes, and I'm starting to feel, okay, maybe, maybe I need to look into that. And, and eventually I'm deciding that I cannot— I, I can't ignore it anymore. Like, if I'm not going to be a part of this change, it's going to happen above my head, okay? I don't want to be affected by it, I want to control it. I want to be a part of it, I want to be the change. And that's the, the story we're going to give. First of all, I want you to put it in the head, like This story is from a year ago, okay? And that's— today I think there's convictions for a lot of things we didn't have there. So basically it looks a lot, but I will talk to you about small AI demo that I built, like the story, and then it's— and how it sparked like a huge change in Monday.com. I'll tell you how you can spot the right opportunities for disruption in AI and how can you maybe change your own organization And lastly, I'm going to give you your own cookbook on how can you disrupt your own company when you go back home, because if there's no lessons, you'll give me bad feedback later. So let's start. This is monday.com, okay? Basically, or at least what it used to be, it's a task management. It's very, very flexible. People build their tasks or leads in a CRM or tickets or anything. Today, spoiler alert, if you open monday.com, it's agents that also use the board. And this is like vibe coding. In this example, it's Base44. It's like a lovable competitor, right? A year ago, it was all the rage. You write which website, which app you want, and it writes code using AI, and you get a full website. And some— about a year ago, I got a message. This is from Eran Rosen on Slack. Eran Rosen. Is leading what we call Monday for Nonprofit. Basically we give Monday for free or nearly for free for nonprofit organizations. There's like a team now in Venezuela working with the earthquake and he told me, look, Base44, one of those vibecoding companies, they're doing a hackathon. They're saying, can you build something for good, like help an NGO with something and we'll give you like full credits and anything. We have a lot of NGOs working with us. Can we maybe do something together? And I said, okay, that sounds interesting, let's meet one of them. And I met with one of them and they had a problem. They were managing donors, right? They have a huge list of donors and they have their details and they need to reach out every couple of months and telling them, okay, look what we did, or would you want to renew your donation? And Monday is basically, because it's very, very generic, You can manage anything with it. The downside of it is that it's basically you now open a donor and it looks like a database, right? Name, email, text, like image is an— or in this case it was even a link, right? It didn't fit like what they need and like they wanted a real donor card, right? With instead of a link, here is an image. Like donation sum, right? With AI suggestions to tell you, okay, this is, you should reach out now. Okay, but now we had a problem, right? We had an enterprise-grade solution. That's a product that is secure, that you can, like, thousands of people can work on it from scale and permission and legal and everything works. And they have like something that they built in 5 minutes and that looks really, really great that doesn't really work in scale and they cannot really use it with sensitive information. Now, in the past, I worked in Monday On one of the places that I touched is there, we called it like a developer platform. Basically, we allowed developers to add plugins to the system. And it was super, super complicated. You had to know how to code and how to code to Monday and like configure like 1,000 different stages. But basically, I hacked it together and I took the page that I generated with Base64 and turned it into the system. So every time that you open a donor, you see their own information. And that way I went to one of those NGOs and said, okay, this is what you wanted, right? Different way to visualize the data. It actually reads it from the system. It actually works. And they really loved it. And I said, okay, what do I do with it? The obvious thing. I recorded a 2-minute video and I sent it on LinkedIn. Look, I did something cool with AI. Praise me. And we're actually going to see this video. Okay, it's going to be— we're going to come back to it. And that's actually the second video. The one— the first one was even worse, and I added subtitles for you. Hi everyone, my name is Amichai Ben-Chen. I want to show you another cool POC of integrating Base64 into monday.com. So let's start with our example here. We'll continue with the NGO theme. So I have a donors board. And I have here all of the active donors or waiting donors. And Monday we have an item card that's showing us all the details from the item. But this is not an item, right? It's a donor. I want to see it in a better way. I want to see the image, not the URL. So what do I do? So I went into Base64, started by just copy and pasting how it looks right now, and little by little I built my donor card. And then I built it, I embedded it as an item view. And here you can see this item takes data from the item view, take from the item. So here is Lorenzo. And if I open here Amichai, you will see that the data will be updated and I see Amichai and it actually shows me the picture and it creates dynamically create AI action item suggestions. So how did I do it? How did I take a Base64 app and then embed it in Monday with dynamic data? So I used monday.com's app framework, right? It's a way to create extensions of Monday, and I basically hosted my Base64 app, my Base64 new view here as an app.

And in Monday, I could just— it now became one of the existing views. So here, Base44 donor card, I edit it and I create as many of those item cards as I want in any format that I want. And I can do the same thing if I want to create another view to show differently the data from a board or create here on the left pane a whole new Base44 application with whatever I want. And I really think there's a great possibility here of extending monday.com with Base44 and no-code apps. I hope it gave you some inspiration and ideas for other apps you can build yourself. Thank you.

Anyway, I posted it on LinkedIn and it blew up. Basically, the founder of Base44 shared it, saying, look what cool things we can build with SaaS. Partners of Monday said, how did you build it? We want to pay you in order for you to build it for customers. Customers started sending me other people in Monday telling me, how the hell did you do that? And it really is like tons of hacking without any code, right? Everything with— I just had like a very, very specific knowledge of how to do 3 different things and I just made them all together. So I did like the only logical thing and I created another video. This time I took like people really, really wanted like a portal in Monday that instead of a board, take the data but turn each one of them into a card with a link and you click on it, it will open and now you have like a service portal or a knowledge base or whatever. People really, really loved it. And of course another one. So in this case, we like, we have a CRM and people like in Monday, we have dashboards that are like you can add the building blocks and different widgets. But it was super hard to add new widgets. It was spaghetti code and nobody wanted it. And it doesn't matter how many of those we are going to build, people always want something else. So I took something that we don't have in Monday, like a Sankey diagram. Those are like all of my leads, contacted leads, and qualified, unqualified, lost, negotiated. And I worked really hard on it so it will look even better than the built-in ones, like the same colors but Even like it was really hard to make an AI of a year and a half ago do this shit. But I sent it and then the CRM people in Monday suddenly, how the hell did you build it? It's like we wanted it for years. And basically every time I'm building another one and I'm creating like my LinkedIn basically at this space was like videos, videos of me taking things from Monday's backlog and just building them as a product manager on the side. And at some point something starts to change. Okay, I do those fun things, but people in Monday starting to say, okay, we can't keep it like that. Like, how can he build something that we are asking for years and can't do and we never be able to build and he builds it like in a cave? And we'll go back to what he did to Monday, but I want to tell you, now I'm giving you like when I'm looking at it back and I'm trying to understand why did it resonate so much? Why did those videos really change people's mind? I want to give you a look at the way that I think about it. And of course, like a product manager, it has to be in a matrix. If it's not in a matrix, it's boring and it's not smart. But this is like the basic effort versus impact. Every time that you have a feature, you ask how hard it is and how valuable it is. Now, most of the things that we build, or that it's really, really high impact and easy effort, those are the quick wins. Build 5 of those in the next sprint, people just love it. You just release them. They might be really, really important, but you just— they're easy to build. And then we have the things that are very impactful and very hard to build. Those are maybe the differentiators. When you look at 2 companies that does the same thing, but the product is really different, that's because you can't build the whole thing. You have to make those hard decisions here. And then you decide, okay, those are the most important things. This is my roadmap for the next couple of quarters. But then there's this place, it's like that graveyard of features that will never be prioritized. Like people want them. Every customer call, they ask for it again and again. And they want you to build it. And like it's obvious that they're valuable. But they're super hard to build and they're not on your roadmap because you made the hard decision as a product manager, right? I'm going to build the most valuable ones. And this is where customers get angry and this is where like sales say product never build the right things. This one. Now, what did I do? I literally looked for all the features in this box. We actually have like a community page when you can vote. Customers can like say, I want this feature, I want a new widget, I want the formulas within dashboards, right? And then we actually tell them, okay, this is coming soon and this is not on roadmap. Like, we actually tell them this is not coming, find like another solution. And I went one by one And you can see widgets was one of the biggest ones repeating. And I literally took them and built. Someone wanted a portal, built a portal. Somebody wanted a widget, built a widget. Because I was looking for the problems, sorry, that are super, super valuable and we don't build. Because what I actually showed is that with this effect of AI, and there's a lot of things that changed, but a lot of the things that were never been prioritized are more quick wins. I'll just give you an example, okay? I worked on Monday automations. It's like if this then that. If a task is done, send an email to someone. And then we have a problem that people are building loops, right? If a task is done, change the status to done. And if the status is done, change the like. And then or sometimes those loop like three automations that are calling each other. And then you have like millions of runs. The whole system is falling. People hate you. And I came to my developer and said, okay. Let's fix the loop problem. And then he told me, look, I have like 3 degrees in mathematics. Loop detection is impossible in production environments. It's like an unsolved problem, whatever. Today, we send like literally the text of the automations to GPT and ask it, is there a loop here? And we get yes, no. Solved 99% of the problem. And like, this is a small example, but we have tons of those. Okay, my developers today, they build features 10 times faster than yesterday. Than a year, than a year for last year. Literally, we can build things that were impossible yesterday are super easy today. And if we don't understand them, this means that the AI revolution is about to happen and we're going to lose. Because if my competitor is understanding that and take all of the differentiators and build all of them now one after the other, my product is going to be yesterday. This— those are the products that fail. And Now I'm telling you, I didn't get it. I did like small videos of cool things and I was looking what is going to annoy people the most. I'm going to build that. But looking back, this really helped people like drop something in their head and understand that they have to change and they have to think differently. And when we think about lessons, we'll go back to that. You can also probably have those kind of features in your company. Okay, so we talked a little bit about my demo and we saw my lovely video. And we talked a little bit about the theory. Now let's go back to the organization because I'm not the only smart person in Monday. Maybe I'm the most annoying or the one best with videos. But other people are also starting to see it. They see my videos but they see the news and they see competitors and they see like O3 comes out and everything changes. And the CEOs and CPOs and VPs, they're saying, okay, we have to change something. We are still building products the old way and we're building the old products. We have to change something. And decided to go on something a bit radical. We called it the Builders AI Month. Builders— now builders is cool, but we were builders before. Like builders, we called everyone in Monday that is building. So it's product and design and analysts and developers and everyone that is basically not sales and marketing. But we took all of them and we told them, look, for one month now you stop working on— stop working on whatever you did. Like, you're not building whatever you decided to do. You're not working on your roadmap. You are going to learn how to use AI and how to build AI and you are going to think about AI. It took more than a month. It was at least 2 in this case, but we had like 3 very big rules. Okay, one, it's not a trade-off with the roadmap. Okay, we don't want AI dust. AI dust is when you say, okay, I have a product and there's a chat. You can talk to it. Nobody use the chat. It's like nice to have. No, take the biggest problem that your customer has, the ones that you already wanted to solve, and now look over a different approach to solve it using AI. We don't want like nice-to-haves. We aim for moonshots. Like we built 4 different— like everything changed something, but we, we built 4 different products that each one of them, if they are going to succeed, they can change the bottom line of the company, right? They're not about— we're not about like making it pretty or using a little bit AI somewhere. And it means— and I'm telling you with the hindsight— some of them failed. Like, we, we launched 3 products, 2 of them were killed, one of them was pivoted, one of them was a big success. Lastly, don't be afraid of potential waste. We literally told teams, look, It's a lot cheaper to build now. We are at startup mode. If 3 teams are going to build the same infra for 3D in a different way, don't care. We don't care about it. Just try to ship something. Just build it. If it will win, we will work on how we consolidate the infras and make it work. Okay? And it was the— when you try to build something new, that's the way. And also when it's so easier to build, that's the way. Basically, they told us We have your backs, right? You can fail, you can go big. There were some developers that we'll talk about in a minute that they weren't productive in this month. The only thing they do is learn how to use Cursor and Cloud Code and like stop writing code and start using AI to do it for them. Okay, today it's even obvious, but a year ago people really, really afraid, afraid of doing it, and some places it is. So again, we gave Cursor and Cloud Code to every developer and every product and every analyst, every designer. We gave everyone like tons of, we like teach them how to do it and people shared everything that they built. We had a competition of who is going to build like the new AI tool that, it's funny but the first place is like tickets to London. But so like, and you don't get it, like the designers started creating PRs of design changes to production. And some developers took a troubleshooting process that took 6 hours and they just automated it with an agent that just reads all the logs and tells you this is the problem. And some products, builds, skills, everyone did something internally or externally. And every week we had a demo. I did a vibecoding workshop. We talked about security with AI. We talked about designing with AI. We talked about PMing with AI. We got people from the inside and the outside. We were a big company, so we had a lot of people that had some background, but to be honest, with the AI, nobody knows. A lot of those professionals or experts, they learned it in the last 3 months. So we literally, everyone took something and they shared it. We had live demos every week and we told, just show me something works. It can work in your own local environment, staging, whatever. You can break it later, but show me something. And people say, okay, I see what you built. We can do it in our product. Like, the point was we are going to do a shift. We are not finishing this month and coming back to the same thing and did like, okay, we did like an AI experiment. We're building something else. Okay, I'll go back to it later, but I want to tell you a little bit about what I did this month because basically they took like the leadership in Monday saw my videos And they asked me, okay, what do you need? What do we need in order to make it real? Because I built it, okay, and it worked, but it worked for one customer when I'm holding my hands and like nobody is moving too much. How do we make it something we can sell to thousands of people or companies or tens of thousands or hundreds of thousands around the world and it will work on scale? What do you need? Basically, I took like 3 developers and a designer. We took— and took a bit more than this month, but Not more, not a lot more than two months, and we built a product that does what I did but automatically without knowing anything. And even worse, there was another team that got a task, and it was the marketing team, and their task was to take my videos and make them real customer-grade videos. So let's see what happened. I'm Amichai. I'm a senior product manager at Monday.com. Let me show you something really crazy that we just built. We took Monday.com. And we took vibe coding to create something bigger than both of them. It's called Monday Vibe, and it will change the way you build business apps. If you don't know what vibe coding is, it's when you tell an AI what product you want and it just builds it. You don't need to know how to code. You don't need to be an engineer. Everyone is a product manager. And I'm saying that as the product manager who built this tool. Monday is already super flexible, but now No product can cover every use case and every need. This was true until Monday Vibe showed up. This is exactly what Monday Vibe does. Just tell it what you need and it builds the product for you. Let me give you some examples of the stuff you can build with it. Let's say you're an event manager and you want to allow people to see the agenda and register. Just tell Vibe and your own interactive event page is live. You want to gather business intelligence on your competitors? We won't tell. Just ask Vibe to build an app that uses AI to create a constantly updating report. You know exactly what they do. Want to create a mobile app that tracks your deliveries? Done! And it's already connected to everything in Monday— boards, dashboards, automations, you name it. I know there are many concerns about permissions, safety, privacy. You don't have to worry about it with Monday Vibe because it's built On monday.com's secured enterprise-grade infrastructure. Don't build it, vibe it. Sorry for making you see an ad. No, but for me, basically taking those small, like, niche videos that I did for fun and making them something real was a, was a crazy experiment. And, um, the— and basically This is from the quarterly results. Right after that, Vibe is the fastest product to surpass $1 million in ARR in monday.com's history. And this is from about 2 quarters ago. They don't allow me to share the new numbers, but they're a lot higher. We basically opened the feature. We added like a small heart icon somewhere. We didn't tell people, we didn't have like a tips scene, nothing. People just discovered it. and start building. Like I see in the logs, in the middle of the night someone sends like the 100th prompt of, no, move my button a little bit to the left, I want it better. Like people were in love with it and broke it and they took basically, I saw what they're building. A lot of them took all of the backlog that they already asked for us and instead of asking us to build it, they just built it. They just did whatever they want and a lot of things, I just met a company, it's a— What's the name? Now a side story. I met a company, it's an energy projects company in the US, and they wanted to get a better feel of what is the status of all the projects. So someone there just built a Polymarket, internal Polymarket, when you can go to every project and bet whether it's going to finish on time and on budget. And you literally can see all the company projects and what people really, really think is going to be the results. They build crazy, crazy things. We would never build a Polymarket for business, even though great startup idea if someone wants, and they just do it. Here's just a more relevant example. This is like, this is a company was paying $100,000 a year for an internal portal for their companies. The company, they literally took the pictures, gave it to Vibe, They got the same thing, $10 a month. And, um, okay, I'm not going to bore you with the details, but it's very, very technical and fun. If you want, we can do it in the Q&A. And so that was my corner, and basically it was like one small hackathon inside 700 different people participating in this AI month, right? We had 71 live demos of people showing here is a new feature, there's a new idea, and there's a new product. 118 captain logs. Captain logs is like every time that you change something, here is a new feature, there's a new change, there's an improvement that actually goes to production. We send the captain logs. 17 workshops, 22 different speakers. And now again, this already— I need to update this slide, but basically we had at least— we have 4 different bets. Like we had Sidekick. Sidekick, it's like ChatGPT only in Monday, and it can read and have all of your context and all of your meetings and all of your data. We had Vibe, which is my product with the vibe coding. We had Magic, which is like configuring your account from scratch. Now we have agents. Basically, this is a Monday today is like 90% agents do the work with you. You can tag them of a task and tell them, just tell me when it works. You can talk to them on WhatsApp, like a completely different thing. And those, like here in, in, in about a couple of months, we got, for example, 67 million AI actions. And when we call an AI action is literally doing something instead of a person. I'll give you an example. You open a ticket and in Monday you could open a ticket and then somebody will read it and then you will say, okay, no, it's not an IT ticket, it's a workspace template ticket because it's physically about the computer and not about the software inside or something. And then it will go to another, like you will wait 8 hours until your ticket gets to the right place and a lot of people will need to read it. Now you open it and I read it and say, okay, that's like another team. By the way, have it in the documentation. Here is the result. Just I'm giving you the answer and closing. Okay, if you have a problem, I'll escalate. Bam, done. Instead of you waiting six hours, instead of having someone in IT like read it and move it around, and it and I close it. And this is like one AI actions, and we're talking in millions. Basically, and more than 50 accounts of Monday, like Monday customers with more than 250,000 customers actually use those. So like the month that we did and the change that we did, we didn't see it at the end of the month, but every month after that and every quarter we see the change in how people work in, and in the product itself. Okay, we're getting to the end. What do you do? How do— what, how— what is your takeaways? So I'll give you like a playbook, okay? We gave a lot of small tidbits, but basically first thing that I want to share with you, don't wait for permissions. Like, I wasn't an AI product manager. I don't think there was a role like that. I, I, I worked on Monday permission in this case, right? I was not even close to whatever this was. Someone asked me for help, asked me if I have like an idea of how to connect Monday to something, and I did it as a favor and it was interesting. You don't need to be the person that is going to revolutionize the product or is the director of the product or AI is in the title. You literally can— it's so easy today to build a prototype or to build a skill or to automate something or anything. Like, so many things are so easy, just do it. Okay, secondly, and that's again, that's like my style, but experiment in public. Don't wait for it to be to be perfect, right? Share it with people. Most people should share it probably inside the company. I shared it outside. I shared it on LinkedIn and to be honest it got to the CEO faster than if I would send it to him directly. And like there is a power in that. You get feedback super fast. Other people tell you, okay, I have an idea, I have a variation of what you shared and I think we can change that. Like other people in your company get the confidence to also share. There's real power here, and again, it stems because it's so easy to build something visual and explainable. Today I, I have a— today I'm doing everything visually with videos. I have like skills, I, I just give it like— I have a customer call, I just send the call to Claude and telling it create a testimonial from it, and it creates like a minute and a half of praises from the customer. It takes from the full hour of text, and it actually takes like different parts of the video when he shows the product and I have something automatically. Super visual, super easy, super easy to share. Same with customer feedback, super easy to share like that. And lastly, people tell, okay, I'm not waiting for permissions, I want to build something, but what the hell should I build? And what I want to say is that the next big idea is not going to be from, like, coming from nowhere. Like, you already have the next big problem at least In your backlogs. People have the same problems. If you are trying to solve something completely new, some people do it, okay, there are some startups, but most of the time it means that you're doing something wrong. You don't want to like invent, here is something completely new, here is a chat that you don't really need. Go to the biggest thing that you already want to build or the biggest thing that your team always wanted to build, the customers are asking and you're not building, and try to ask, okay, maybe today it's easy. Like maybe something that was impossible yesterday or super expensive, or the only way to do it was like with 600— with 10 people for 6 months. And basically we can build a version of it like a couple of weeks now with AI. That's the next big idea. You don't need to look far. So go back home, look for the best thing that you're not building, ask yourself, can I build it? And just build something and, and then share it. That's the way to get 700 people behind you also changing the way that they think. So thank you. This is my LinkedIn. I still share videos and ideas and cool stuff. It worked once, so I'm— maybe it will work the second time. I'd like to connect. And thank you. Thank you very much. Thank you. Don't. Okay, so I think the question was how do you, like, in my specific example, the power was that it's enterprise-grade and secure, but every time that you demo something, it's usually the opposite of secure. And what I'm trying to say, there's a difference between getting the leverage or the resources you need in order to build something and actually building it. Okay? If I would go to my company that are not doing vibe coding, none of the competitors are doing vibe coding, and I would tell them, I want to build a new product category, give me like your best developers, tons of money, and I'll just build it. They will tell me— and I'm a product manager working on permissions— they will tell me, who are you and what do you want from us? Here's the door. But when I build a POC that is not secured, it's not scaling, it's worth nothing. But it shows here is a real product, a customer problem that you really feel, you can see it, it's really, really clear. And here is an outline of a solution. Look, it's so possible that I literally showed you one, like I built one in like an hour and I'm not a developer and I was able to build an MVP. or a POC for it. Now the chance that they need to take in order to give me the resources I need to build it for real, like the risk is so low because now they are sold on the customer problem. They are sold that technologically we can build it because I built a POC, okay? And they are sold that like they can even trust me because I have the judgment to take it into a 2-minute video and just tell you here is a customer, Here is the problem, here is the solution, here is the scale. And like it was so easy to digest, they say, okay. And again, I needed like 15 of those, but this is what makes it like, this would give you the ability to build now an enterprise grade. And even to give you, to be honest, the first version of this product didn't really work. It worked like, I don't know, 1 of 3 times. You would give a prompt, it will break, maybe the second, maybe it will work. Like customers would— but even though it was unstable and it didn't work and it was super ugly, people like— the problem was so clear that people will just give it the 5th prompt until it will work. Okay, so, and that happened only because I invested so much time in explaining what the problem is and showing the vision that people were willing to use the worst product ever for a long time In order to get there. Today, even today, it's an AI product, right? Once in a while it will destroy everything and it can't, we didn't give it the ability to delete anything, but you will get like something really, really bad and ugly and you will say, I hate this product. And we're losing like some percent of customers because this is what the AI gave them on the first impression. But it's about, again, the conviction and the customer problem that you can do when you build in public. Yeah, so there are 2. I will answer not only on Vibe but also on monday.com. So Monday changed the product completely. We basically took everything that we built in the last 10 years, it's there, but it's now a backdrop to an agentic product. Like we said, until today we have a table and it, it was where you manage your work. Like, you go somewhere and you finish a task, and you go to Monday and change the status to done, and now everyone can see the status and we work together. Now we say our winning picture is not to show you what is the status, is to do the work for you. Like, we want that more and more time— like, task of your day will be done by Monday. And Monday you'll be able to communicate with the agent and see like what is the status of everything and maybe approve things that need your So first of all, like, we— half of the people stopped doing what they worked on and they are now working on agents and they'll stop building like that. I think everyone who works on dashboards and widgets, we don't need them anymore. Like, vibe code it, okay? We add another package to the AI, it will have like 1,000 more widgets. We don't need a team that builds widgets, okay? What did they build? Like an agentic product. That's on the, on the like strategy side. If you open the— and it was losing at the beginning, right? You would— we would see the drop. We have like a sign-up and people knew what Monday is, and when they, they were looking for task management and they got an agentic product, say what it is, it's not what I want, right? It's actually losing the habitus, and we are now like improving the whole funnel to sell something else. And in the building side, it is 10 times faster to build stuff like when everything is agentic. Like today, my developers, they're building the product like 80%, and then they give it to the designer to make the— make it look real, like, like the Figma, and a lot better. Like, some features don't have a Figma. Like, the, the developer gets the general idea, all the backend, and the designer is just having a PR with a solution. And we take like 10, 10 back and forth for being closed. Or like, I sometimes— we have a— like, I have a bug, I tag an agent in Slack, and he says, okay, here is a PR with a solution to the bug. Right, we can build 10 times faster. Now the question is, what should you build and how do you take it to market and how do you make customers use it? Like all the rest of the problems that we didn't solve yet. But just by the way, like, again, something that I had a video of, it started from a video of mine. If you say, if you put your details in the Monday website, in 15 seconds an agent calls you and say, okay, I saw that you put your details in the website. I'd like to hear about your company and use case, and then I'd love to set a meeting with one of our reps. Before that, it was maybe taking 3 days, the lead was called, nobody remembers that they even go to your website, and half of the people we talk to is like a 5-people company that we don't have the resources to answer. Now everyone gets a call immediately, they're still on the website, they get it, the agent even funny, like people love to talk to it, and we automatically say, okay, if it's under 100 people in the company, it's out of our radar, or if it's a bad use case and we just tell them, sorry, we can't. Or, and if it's like a huge one, it's escalated, then in 5 minutes they get like a meeting with someone. And like even how we do business is changing. And so every part of the company is like asking, what am I doing that shouldn't be done with a human? And even if the technology is not 100% there, we're changing that. Like we have a bot that like query the database. I like— everyone is an analyst You can literally write in English, how many people started using Vibe in the last 24 hours, and you get a graph back. Reads the whole thing. And like, people who would never touch SQL can just do it. Yeah, so the question was like, do we look at what people build and then incorporate in the product? So I will answer a couple of them. We can't see what people build. We look at the proprietary data, it's like it can have, people can put in the title of the app something private or something about their company that they cannot share. So legally I cannot see anything anyone built unless they share it with me. Having said that, we are looking at what people are building and we are looking at what are the top use cases and we are saying, What are the top use cases? but with something that is a bit more anonymized. And we do say, okay, those are the top use cases and we might build them. Just for example, a lot of people built, they took the board and they turned it into a PowerPoint. Literally a live PowerPoint, like a presentation that you, because it's code, it doesn't have to be a dashboard. It could just be like a deck with slides that you can slide. Then we said, okay, why don't we like have a button of just turn it into slides? That use Vi behind the scene, but So we sometimes package it or we have templates. Start with something that we built, then modify it. Now we have parts of the product, they literally have some of the part is hardcoded with an old developer and some of the part of the product is, okay, we started it for you but you can change it any way that you want. And some of it, okay, extend it, put your own. So yes, we are looking what people, if something is really easy to build with vibecoding, as I said, We just give it for vibe coding. And if it's— sometimes we do want to take some part of the responsibility more on the experience. So we have an experience that you start with, and then you can still modify. And if you don't like it, say, like, bring it back to, like, the default. So we are looking at different hybrid ways to use the new UI elements. What are some of the challenges you're facing? month. I will say that our company is a bit extreme. We basically told everyone, if your productivity is zero in this month but you learned how to use AI tools, and from now on you know how to write code with Kelson Cloud Code, it's a month well spent. That's like, at least on the personal level. On the, on the group level, like, every, every team and every group and every, like, the director in Monday said, like, this is goal for this. Like, here is the biggest bet that we are working on, here are the smallest bets. And like, some teams like build something like just built an API or MCP. Some teams said, okay, what we can do in this month is take whatever it is that you can only do with the UI today, make sure that every agent can do. And like half of the— like nearly every team had that to just open up things that were impossible, and now every AI agent or a feature can use it. So some teams did something that are smaller if they didn't have the vision or the resources, something huge. And the things that basically we did was or there's an incident now and we don't, like support and stuff like that we didn't lose, or something that we had like a real commitment that we cannot shake no matter what and we gave it to a huge customer, okay, so we did that. But unless you had a very good reason, Everyone was on it, but I'm just saying this is our goal. Like, this is how the company works. This year we had like an Agentic Week and we had like a $15,000 first prize for the one who will build the best agent that is going to save time for themselves or customers or whatever. And people built like 3,000 different agents and about 20% of them are still alive, which is pretty cool. That's amazing. Pretty much pretty big for this number. And like every— everyone in Monday, not only builders, like really people who are technophobes, built an agent with their own hand in Monday's tools and used it on something. Because like that's our way to— like we look at it, we are in startup mode. Like it's a company that does more than a billion dollars in ARR, right? But we look at not what we already did and we're saying if we are not going to do something drastic Like, we are going to lose. Cloud Code or any other startup. So that's the way that we see it. Again, we are, to be honest, we're an Israeli company and our founders are a bit crazy and we do things really, really extreme. But like, we believe— like, we— I was— I'm a part of it, right? But we do changes drastically, not bit by bit. Otherwise, you will never get there. And to be honest, some of the things we did in this month was were bad, but we do now things today that were magic a year ago because people like really shifted their mind back then. Look, the way that I see it, we don't look at it as a hackathon. We also have hackathon. Hackathon is like we now need to make sure that all the AI products are using the same connected to the same payments mechanism and like let's get a developer from every team, give them one day, everyone in the same room, just finish it. That's a hackathon. This was, it was a month that cost 2 months, right? And it was like the people, like the CTO and the CEOs and every VP was a part of it. Like think about it as a hackathon that has, to be honest, a lot more money than any hackathon you can think of and it has the focus of the people that what they do is run a company. For their main job and they did it for the last 15 years in this specific company and you have 700 different people doing it, like, that's a lot more. And secondly, we do believe in transformation. So as I said, we had like the Agent Week, which was not builder-focused. It was basically, we said, okay, builders know. Like developers, they use Claude code, they will be okay. What about the sales people? What about the implementers? What about the marketing? What about finance? What about legal? So we did the Agent Week and we said everyone, like this one, 3,500 people, everyone is going to build an agent and we're going to have open hours when someone can teach you and help you. We're going to have recordings, we're going to have sessions, and we're going to have a competition with real money that's worth your time until everyone will be actually on it and do something. And like this, it took a full week and it's a week that we didn't do other things, but it was directly connected to the company strategy. Every sale or marketing or that person that built an agent and then need to then market or sell it now built with their own hands and used the product. There was a real outcome and there was a real reason and people wanted to be in it and we had like real goals and real incentives, so it worked. If we would just tell people, okay, and we had those in the past, like, okay, build something. You have a day off, build something. You have a hackathon. And in the end, we will take the first one and we'll give them like a small prize. Then most people will find a way to get out. But if it's a full week, all the television are talking about the Agent Week, there's a huge prize, people are asking you, okay, I saw that you do something, like, are you— do you want to take a chance to get this TV? $5,000, I don't know, $15,000, something was crazy, I don't remember. But like this makes people be really a part of it. So you need, it needs to be aligned with the business, you need leadership to be a part of it, it needs to take time, you need to plan for it, and then it can really make a shift. If it's a hackathon, it's more of a, you know, it's a fun day for employees, it's not a mechanism.