Building AI Automations Just Changed Forever

So, you've heard of Vibe coding, maybe Vibe Marketing as well. Well, today I introduce you to Vibe Automation, the ability to create entire AI automation workflows from a single text prop. We are now at this point where we're seeing platforms like any 10 integrate these features into their platform. There's never been a better time to bring on someone like Mark Casher, a good friend of mine and member of my AAA accelerator program to break down what Vibe Automation actually is, what it means for beginners trying to get in and create AI automation businesses and also give us a demo to see if this really delivers on the promise of being able to get that instant workflow and also breaking down the ramifications of this for existing AI automation agencies. Tons to talk about here. An exciting new frontier of the AI automation space has opened up. I hope you enjoy. Well, good

to see you and I'm excited to dig in automation stuff because um it's definitely an important week to talk about it with all of the new stuff coming out. So, um let's get into it. — So, when we go into this new golden era, this Vibe Automation Frenzy era right now, we have basically tons of top companies and brand new entrance trying to create this text to workflow experience. So we've had text to image, text, text to video, and now pretty much across the board, everyone's moving to text to workflow. So where you were like you had to learn all these nodes for every single thing like make. com or edit knowing the differences between them or basically going into a place where all you have to do is imagine what the workflow looks like and then you can worry about the nuts and bolts like the last 20% once it's on screen which is a big leap from where we've been for the last couple years. Obviously, we both are aware of the new entrant here from OpenAI that just dropped this week, the agent builder. For me, at least I find a little bit underwhelming, but I think we both know they're going to want to compete on the same level. So, it'll be a matter of time till they're in this game as well. So if we go back to you know last couple years we've had two main competitors in the automation space and then a eventually got added last year once they got uh the agent node and everyone went crazy about that but it's been the zap year and make game for a long time and as soon as this AI agent wave was born last year you can literally see if you look at the Google trend chart for NN like this is approximately when people started making videos on it especially people like Nate and other people creating workflows with that agent node and it just went hockey stick after that. So then we got nen into the market with very developer focused but the only thing with nen was it came with a pretty steep learning curve. So make. com nice, bubbly, pretty, but where it was annoying was you had to come up with five or six nodes to do one action that you could do in like Python or string or anything like that. So you had to like maggyver your way around it. NN gave you flexibility and it was way more potent, but it came with this consequence of you having to learn kind of like a new dialect of automation and especially like going from zero to one with NAND was a lot harder than going with make. com again where you have that visual stimuli and feedback. So you basically had to make a choice for a while. Do you want robust and like more scalable and cheaper or do you want pretty h holds your hand intuitive but might cost a little bit more? So you actually had to make a binary decision for the most part. — Yeah. And they had you because of that agent step, right? Everyone wanted to use that. Make didn't have a good alternative and in my opinion still doesn't really have a very good alternative to it. Zapia certainly didn't. So they had the whole game in their hands with that agent step and people were forced to go through those hurdles um just to get the that kind of output that they wanted. — Exactly. Yeah. So you had the shiny thing waiting for you but to use it to harness its power you had to go through the mud. So eventually I ended up asking this question because I was a make guy kind of like you and I was just asking myself is it worth it me learning this brand new thing. So the question was what if I could just describe what I wanted? That was the main premise. So initially I wanted this like we had a workflow. Basically the workflow looked like this, right? So you have a web hook and you have some tools and you have an AI agent. So it didn't seem like rocket science. U but in behind the scenes this workflow ended up looking like this where it's JSON which stands for JavaScript object notation. — But that's the whole thing. — Yeah, that's literally the whole thing. And if you look for the agent node, — that's like so simple compared to like a uh a make. com version, you know, like the make. com um blueprints and the export are like way bigger than that. — Yeah. And like this this little baby here, this lang chain thing is what the AI agent is. So once I saw this, I'm like, okay, so JSON is something that a language model can output. And now we're in a place, at least back in January, February, where we could output way more tokens than we used to. So what if I could just generate the JSON that represented this. That was basically the main premise for this question. And then this first approach was born where I just used chat GBT and at the time deep research to go and look through the NEN GitHub because Naden unlike make. com they've open sourced a lot of the code of how the platform works. So I'm like, okay, what if we had a language model? Go and reference that and then go and reference their documentation look for a couple examples of workflows and come back with a full report. And initially we got things like this where there was things on the workflow, but there's things broken like there's some like non-existent nodes. It was imperfect and sometimes it was even more work to sometimes do this versus just starting from zero. So it was okay. It kind of like validated the concept. But I wanted to go to the next step which is like how do we make this a lot more tangible and then tried out Claude and for some reason obviously Claude is a better coding model. Obviously depending on who you ask on what day for the most part it's been a better coding model. It was way better dealing with the JSON and kind of understanding the main bedrock foundation of how to construct it. So you'd have less occurrences even with like claude 3. 7 of completely broken workflows. Like you'd have something on screen that at least would give you an 8020. And that's where these bad boys were born that both happen to be my only 100k plus videos on the channel. — Yeah. These are top performers, right? These things blew up because I think a lot of people were in the exact same position as you trying to make that jump from maybe make. com over to uh over to Nad. I mean Yeah. I mean I remember watching these and I paid for your like Gumroad thing. I got your pack and I was trying it out myself and I was like, man, this is freaking awesome. I think more people need to see this. — Yeah. And then all of a sudden I started seeing the same title like 10 times. — Yeah, the copycaters. Yeah, they got you clicked to that. That's how you know you've made it in the space when your videos are getting copycated like immediately. So, congrats, bro. — I appreciate it, man. It's all good. — You're influential now. So in terms of the workflow if initially we just had the endto-end workflow and then I tacked on okay what if I gave five examples of workflows that had AI agent nodes and the reason why I did that is Claude had no idea what the agent node was because it was trained as of 2023. So I had to spoon feed it like a cheat sheet of what the new world of N& N looked like. So when I had a clawed project and I dumped in some example workflows and then a really solid prompt, I was now able to crank out remixes of different AI agent workflows. And then I hit another wall which was what happens if it doesn't know that a tool exists like a native tool like ClickUp or Salesforce. And then we added on one more thing which is okay what if I pulled on a canvas in NAN and I added every single tool that I cared about and I downloaded that as a JSON. So now I would have a cheat sheet of how each tool looks like as of now. So again very imperfect and hacky solutions. But the goal was eventually to get it to a place where it was even stronger. And that's where I think a few months ago the world got a really helpful MCP server called the editen MCP and it was started out as a very secret tiny little workflow or GitHub repo rather and then it kind of blew up. I think right now it's in the thousands of people that have starred the uh repo and then this

combined this MCP server with something like cursor and now this had its own cheat sheet of all the nodes the node properties the documentation a bunch of workflows to reference and this gave you even more nitros so this now gave birth to the instantly build agents but with MCPs so it's kind of like been going from manual to copycatting to knowledgebased to MCP with a knowledge space and finally I'd say a lot of these folks started seeing the trend and vibe automation started getting a lot more traction and I think one of the first companies to offer it out of all of them I think was this uh string by pipe. Yeah, pipe dream like came pretty early a few like a month ago or two months ago with the string app and then all of these started bouncing up and then we got Lindy 3 here. Then we got Zap year at least now changing their homepage look like it and then NAND finally released theirs in pretty much alpha. It's still being rolled out. It's I think it's at 5 to 10% of the user base has access to it and I know they're going to release it wide very soon. But now we're in a place where all of these automation vendors and soon probably OpenAI will add the ability to go from like a thought to a workflow really quickly. — That's the future 100%. And I guess the real question now that you people kind of want to know is uh how good are the things that are coming out of it at this point? — Yeah. So I've tried all of them. Now obviously when you go with things like string and lindy you're locked into those platforms. So it's kind of like building within that platform. Whereas something with like nen or zap year technically a lot of those workflows will generalize across tools that offer the same thing like web hook nodes loop all everything around that from any it is really solid at the 8020. It has a native understanding of how to build a workflow which I find to be really cool. It's not just doing what I was doing which is kind of spoon feeding an LLM. here's a cheat sheet, do this and like build it like this. There's some form of rationale in the way it lays out the workflow. You can see it drafting it. Um, and it's aware of its own nodes. The one flaw it has that I found is when they come out with new community nodes, like new integrations, it doesn't necessarily know about it yet. So, you kind of like have to spoon feed it like no, no, like there is a community node for this. Go and pull it on screen. So, I think you there's a little bit of culture. — Do you have access to it? — Yeah, of course I do. Yeah, let's do it. — Let's just chuck something in and give it a go. — No problem. I'm loading this up. Try and think of something you want to ask for — a classification step like uh say we want to do a query of a database and we have two different agents that query different parts of the database like we have the commerce agent and then I think there's like a legal agent or something um where it's got a there was classification split and then each agent gets queried based off uh based off that input. — Gotcha. What was the first one? You said the legal E. Second one. — If you just pull up, we can probably just like snag it off. Um, if you go on to agent kit quickly, we can just try to get it to replicate that. — Agent builder. So, so agent kit is the SDK and then agent builder is the app. — Yeah. — So, yeah, if you go the structured data Q& A on the right. — Yeah, this one. I got you. — Um, — all commerce uh commerce query. I got you. Okay. Yeah. Well, what you can do is here if you want to cheat code this kind of — screenshot it. Yeah. — Hold on to YouTube. — If you're going to coin Vibe Automation, we got to go all the way. — So, let's pull this up. — This is a really good test. I mean, with I'm super excited to see what Opening Eye rolls out in this way um to compete with something like this. — We'll see. Okay. Hey, can you quickly do some research on this new feature from Naden as of October 2025 called text to workflow? There's some YouTube videos out there and look at this workflow. We want to basically emulate it. So come up with a short prompt. Let's say less than 5 600 characters where we can emulate a lot of this design but using any. So now I'm just going to make Claude do the heavy lifting. I'll make sure it says any not something else. Okay, perfect. So, this should go do some research. Pretty quick one. Come back with a prompt. I'll paste that prompt.

And that's where you really get into the vibing, right? So, you're vibing your prompt. You're vibing your workflow. You're going through the whole process with less manual intervention. Here you go. Here's your prompt. All right. And then let's — think it's a bit skinny. — All right. So, click on create workflow. And then this is the cool part — for user queries for safety. — Oh, it's kind of like a Lavable style editor. — Yeah, basically. And then you can see all the nodes pulled up on screen. — All right. So, you can see here it's pulled up the nodes on the canvas. — Step one, — what? Oh, — looks pretty similar. — Top left. — We have a floater. Yeah, you can tell it that there's a floater and it'll remove it. Yeah. And what's cool is if you change something in the prompts, it knows about it because it has access to the canvas and all the JSON behind it. So even if you make changes, it can still interact with you. — So we've got the chat trigger. It's even put a little guardrail thing in there as well. — Yeah, I think it actually used a tool that I'm not familiar with. Safety check. All right. So I'll delete this little thing. But what this is the part that's cool, they don't have an agent builder yet is when you click on any of these, it comes out with a meta prompt. Like there's already prompts in the workflows. — Wow. Okay. And so maybe that integration with the tool um the uh guardrail thing might trip us up here if we tried to run it. — Yeah. This would be I'd have to like knock this out. Yeah. — And then we'd add a little node that says safe. But as soon as I validate all of these here, meaning like I just double click log in to my account. Just double click into that. That's only annoying thing. Doesn't like auto detect you're logged in. So let's double check. Boom. And then we could do like a set node. — Okay. Yeah. Do you want to just No, maybe just delete the workflow uh the query save thing. We could assume that you put in the guard rail if you wanted to, but you can just run it from there. Let's keep — I think there's a recommended one. If you go on to the agent builder, just ask something about like what was the most popular product in 2015 or something. — I gota All right. So, let's do that. What was the most popular product in 2015? So, there we go. — Personnel queries probably not. All right. — Let me see. Boom. — So, it's match matching up all the variables is the — missing my got that from the first part. Yeah. — Yes. So let's do that. I think now if we do this run it back should be good. — It's getting the classification wrong as well. It's calling it a personel query. — I think this is a really good example like this doesn't think for you. You still have to use your brain. It just — specificity in the prompt. Right. — Exactly. So that's a skill issue on my part. — So in here we just go into that in the domain classification agent tweak it. Give it more context. I mean it doesn't really know what the what commerce query or personnel query is. So, how much can you stuff into that uh into the text like input for the text to workflow? But can you pack it out or is there a character limit? Cuz — yeah, there's a character limit. I think it's around like 800 the last time I tried some 8 900. — I mean, if they expand that out to like a few thousand and then they'll add it like allow it to think over that and really like build a plan up on its end, then you can just dump in tons of information about the use case. So, I mean, I'm pretty impressed to be honest. It's a small of the prompt there would mean it's routing up. You tell it. We didn't really tell it exactly what we meant by personnel or commerce query. Um but it seems to have done it right. So — I think that's pretty in team gets a pass on that. What do you think? — Yeah. No, absolutely. And the thing is like also tell you if you're a complete beginner, a noob exactly how to use the workflow and how to set it up and like walking through what's happening. And then most people usually just want to ask a question like what the heck is a domain classification parser? Like what does this do? And you can just ask that in natural language and that's those are the things that six months ago were really hard to go from zero to one but now that's off the table in terms of like that ignorance bias. — Does this mean that uh Nit YouTubers are a dying breed? — I would say it means that selling NAN templates as like your entire business strategy is probably not the most robust thing in the next few months. — Yeah. Interesting. Well, I mean we can open this up to the broader conversation now if you jump back over to the — Yeah. Yeah, — the board. Is this really that much of a game changer? Say if you're a beginner coming into it into the AI automation space, is this the your saving grace? Is this really really going to help you? Um or is there still a lot of fundamentals that you need to learn? Um or I mean vice versa. You could pick them up along the way? — Yeah. No, that's a great question. So, I'd say if someone was entering brand new, this is good in where we're heading as an industry and in terms of being able to sell these for how much a lot of us have been able to sell them up until now because there was no cheat code. We had to build it ourselves or find our own cheat codes. The value of a single automation like the perceived value intrinsically will be lower. But the goal was never just to build the automation, right? The goal was to make companies AI first. So if you were to go to a company now and tell them, listen, we can now really like make your department or these systems more AI first and get you ready for the future, the automation part, the time building it and testing it, that's gone down precipitously. That will continue to go down. Now, you can't just not know what you're doing. Exhibit A of what we just went through right now. If we don't know the business premise, we don't understand what it takes to get there, then you have a co-pilot. You still need to be able to drive the plane. So, it doesn't like absolve you of understanding how this stuff works. But in terms of going through the nitty-gritty, it helps you prevent having to go get in the weeds as much as you did where you'd get stuck on one node for like five hours and have no clue why it's wrong. In terms of like before and after the opportunity like before let's say average workflow assuming it's not nothing too complex could be like an hour two hours for the average workflow assuming it's like what five to 10 nodes it's kind of predictable the data set the data itself is not in a very janky method it would be slower and there would be a lot of iterations but it was really annoying sometimes to deal with clients that had changing scopes or requirements because as those requirements would change, you would have to go back to the drawing board, change the nodes, retest from scratch, and that it whole iteration was painful. But now you could do like technically depending on what the requirement is pretty much same day delivery and you could kind of do bulk like instead of focusing on the individual automation, you can focus on the service which is just empowering the business or the business owner or whoever with the ability to go much faster. Yeah, it's interesting looking at this now because we ran a we ran an

offer called Morningside automation way back where we were just specifically uh offering um I can't remember how many Josh do you remember how many when we did Morningside automation it was like three or four grand per month and you get like three to five automation — five six grand a month — yeah five to I think people have tested it a bit more as well but um where you would have like an ongoing retainer service kind of like the design joy model but for design, but this would be for automation. Um, and we just found that it was the tools just weren't there and it required so much effort. Um, and then also you had scope creep as well where clients would be asking for like kind of full stack stuff and it's like no, we're only really doing automation. So, I think these kind of offers are probably something good for beginners to look into if you want to run more of a retainer based model um and you're just able to come on here and deliver a lot of these workflows a lot faster um than we used to be able to do before. So, it's pretty good overall. And I would say the question is whether or not the price that you can charge for a given automation is only going to necessarily like decrease. One if the business owner is aware that this exists and that you're using it. And then the second one is if you have compared competition in their inbox or in their you know like actually talking to them and making lower bids than you. So and we know this stuff is still what pretty like largely untapped. There's not a huge amount of people who are going to be competing over the same projects. So, you can probably get away with keeping those margins uh that those prices high and have like really juicy margins for a while here. Um, and then over time as you start to notice like those deals not going your way, then you have to start bringing the prices down or making more compelling package. Um, but I still think it's like the golden era, the golden age uh for automation has really begun with this kind of stuff. — Yeah, I would I usually call that naive arbitrage. So, there's a window of time where people don't know that this is that easy. So you can capitalize on that opportunity. If you want to hump hop on that, the other side of that coin for me is right now when I get on paid consults, instead of just like making face noise at each other to try to understand what they're trying to accomplish, I will actually pull up the NN text to workflow and be like, "Oh, is this what you mean? " and then we'll go through the workflow or before you even hop on the call, you can look way more polished and way more professional by coming with a couple workflows and like walking through the workflows as a part of the discovery call instead of having like this imagination of what would be amazing, which most clients do, especially the ones that are less technically savvy. They talk about the blue sky. But if you can like ground the conver conversation with what's actually possible, it might even help with discovery, especially if you come prepared looking polished. They don't know that you haven't wrote written those prompts yourself in the text to workflow. So there is a level of polish arbitrage too that you can take care of. So you can basically vibe automate before you even hop on the call with them. And when you're on the call, you just look a lot more not prestigious, but like you you've come doing your homework, not just on the company, but what you can even offer them. — Yeah. And I think the same sort of thing goes with say lead magnets or if you're doing cold outreach and things like this. Um people are already starting to send me and my team like vibecoded websites that are personalized, right? So it's like a bolt link. Um and they're saying, "Hey, I've made this thing for you. " And it's fairly obvious to us what they've done. But if you could have a say lead magnet on your web page where so they or a lead form where they fill in information about what they're looking for and then the email response to that is literally like I don't know if you can I assume you can't hook in via API to the text workflow thing but if you built if you either had a VA or you had a browser automation where you could like vibe automate the draft of that just the visual of the NAM workflow like hey I just made this for you is this what you were interested in like the chance of closing a deal is going to just like skyrocket. — Yeah. And the mega cheat code is we just did this for one, right? What if all you did was say the name of the company and you did a deep research on whatever platform you want and say, "All right, give me six of these, one for each department and then your VA or your browser automation like you said, you go and draft five automations and you are locked and loaded for that meeting. " — It's almost to pick up on these strats, man. I really think it's a ridiculously edged effort right now. — Yeah, absolutely. And then the next one, this is more so if you're an agency owner. A lot of times people start building before they should really start building. It's not really necessarily on the agency owner, but maybe the client thinks they know what they want to build, but once you're in the weeds, they look at it visibly and you could have spent two, three weeks in the trenches. Then you come back and they say, "Oh, ah, I wish you talked to me before. This is not really how we do it. Jim usually does this. " And then Sarah actually does submits this in the Google form and there's all this new stuff that comes out. It's kind of like vibe coding where if you vibe code the MVP, you didn't pay four developers or a full stack to build that application. They can visually see the feedback loop a lot sooner. So if there are iterations, if scope does need to change, it's not a lot off your back to have to do that. So once you have the skeleton and you guys agree in principle that this is what you're trying to build, now you can invest your time in building the workflow properly, making sure it's robust, it has the guardrails without having to go through that really painful feedback loop where sometimes you make changes for free just because it's kind of awkward to ask for more money just because the scope changed 5% or 2%. — Overall, this just great news like a lot of these dev tools that are coming up for us as agencies, it's such a good time to get in. I remember when we were coming in and I was using texture 3 via the the chat the text completions API to try and build stuff for our initial clients at Morningside and now people can come on with minimal experience and viable to make their way to the first client. Like I think really the big win here is probably for beginners trying to sign their first client where they can look way smarter than they are when the client literally mentions something or the prospect is saying, "Oh yeah, I think something like this would be interesting. " and they can go on either live on a call or they can send it via their inbox or like we just mentioned and look way smarter than they probably are because you've essentially got an AI assistant every step of the way in it. So whether it's going to be Naden, whether it's Zap here or make or maybe OpenAI's agent kit and agent builder, I think they're going to drop one of these soon. It's a it's such a great time to get into this stuff. — Yeah. And pretty much especially when we talk about scale, you obviously own Morningside. Um I run Prompt Advisors. Eventually, when you get to higher scale clients like enterprise, some of these automations, depending on how many times they're running, they might not even be scalable enough for the proper build. But this will allow you to go from zero to something that can operate somewhat like in alpha or even in beta depending on scale. And you can see how it runs where imagine you can now vibe code an app and then the back end is a vibecoded end workflow and then that can run for a bit until they test it out and see the bugs and then once you're sure that they really want to invest in this is exactly what they — vibe marketing it. — You vibe market it. That's it. The whole ecosystem just a vibe ecosystem. — This is the vibe trifecta. Yeah. No, sorry. What were you saying? — No, it is the whole idea of vibing. No, I mean — you have a serious audience to the internet though. — Yeah, I mean I was just going to say building for enterprise involves a lot of custom code, not just like automation tools a lot of times. So if you can prove the concept through vibing, it just lowers your risk and your overhead as an agency. Whereas like I've spent 80 $100,000 in just developer cost sometimes just to go from A to B and have like an MVP for the client. It'll be cheaper for the client, cheaper for you, and just a better experience across the board. — I think with the these shocking numbers we're seeing about like the the AI pilots and how they're going within enterprises, this sort of stuff is really like partially the antidote to that, being able to test things a lot for I mean, in order to have an ROI, if the cost was drastically lower, it's much easier to get an a positive ROI, right? And so that 95% stat is all about these AI pilots that they're doing not delivering the expected ROI. So when the cost comes down a lot, either they're testing internally, it could be good for companies with their own AI automation team or trying to do it internally. Or if they're working with agencies like us, then you've got a much lower cost in many cases. So it should be a lot easier to get a positive ROI. Or you can just test a lot more different use cases um in order to see which one's actually going to provide a valable ROI. So you could get like five systems for the cost of what used to cost one. Um but that's only if we let them get away with the naivity arbitrage. um and give them that highest price. — Yeah. And like where we're heading anyway in the next 12 to 18 months is all kinds of things that historically have been really hard to do will get vibed vibe automated in some way. And what will matter is the strategy consulting like how do you go into a business and identify what needs to be fixed? Are they even eligible have AI in the business? So you can focus on things that move the needle versus all these things that are intricacies or in the weeds kind of workflows. — Yep. No, I think the consulting boom's already kicked off. We're seeing it on our side and know a lot of people are selling these smaller AI audits, but the value of the agency sort of slides towards being able to pick the shots, right? They're coming to you not because oh yes, because they want it built. Um but the building is becoming so easy with these kind of tools. Uh it's the picking of the shots when you've got to prioritize resources and say, "Okay, what's actually going to provide the biggest lift here? " So the strategy around that is really where the money is going. And I also think you've got strategy, then you've got the actual development and setting up, but then for a lot of uh a lot of systems, particularly important ones in the say the acquisition funnel. You've got the agency I believe is going to take up a much bigger role around optimization and management of those systems because if you have major acquisition, say if it's a it's an appointment setter or it's a receptionist or if it's a like a speed to lead system, these things are like really integral parts of different channels or different parts of the funnel. And so if you could just set it up and leave it there, but there's so much more performance you could squeeze out of it. And so agencies are going to shift more towards not only just setting it up, but saying, "Hey, look, we're going to manage it and optimize it over time. " So a lot of the platforms we use don't necessarily have the tracking and the optimization, the testing that we need. So I think on the even on the software side, you're going to start seeing these specialized platforms that allow for highly specialized and optimized versions of these core use cases. Um, which is something I'm excited to see and have play around with. — Yeah. And the last thing that I would say is one thing we experimented a lot with the text to workflow was evolving an existing automation. So let's say you made a lead qualification automation. You can now technically go back to that workflow on any given day and say, you know what, can you extend it? So after they're qualified, we actually have a nurture sequence and you can keep like extending the power of this workflow and over time like you'll be able to build a full system end to end that you build progressively as you prove out each part of the business concept. — Love it. Yeah. I mean adding on voice agents and stuff like that eventually you've got like a lead form trigger and then contract signed at the end, you know, that'd be the dream. — Yeah. — Mark, mate, thank you so much for coming on. uh you are the viable automation man um of YouTube and uh you've really been pushing the space forward. I know you worked uh worked a lot with internet to get these sort of things rolled out. So um massive thanks to you and everything you're doing in the space. You put a lot of great information out there helping a lot of people. So it's been a pleasure to have you on and um yeah look forward to seeing where this stuff goes. If we get a similar thing for the agent kit then you can be sure that we'll jump back on here and do a breakdown. — Right on. Likewise. And yeah, you're the one leading the space. We're just trying to contribute as we could. — If you want to re reach out to uh Mark or watch his YouTube channel, he's got some awesome stuff on there. I'll link all his details down below, his agency and whatnot as well. Um but yeah, see you in the next