The AI agent blueprint that actually grows your business

[0:00] In this AI Creator Workshop session, you learn the AI agent blueprint that actually grows your business. Not just random shiny tools, but the right AI agents built in the right order based on what your business actually needs. We walk you through over 25 real agents across marketing, sales and operations. And I'm super excited to be joined by Maeve Ferguson who builds diagnostic assessment systems and AI agents for some of the world's top thought leaders and experts.

[0:30] With that said, let's dive right into the session. Warm welcome to the session, Maeve. Hey, thank you so much for having me. Yeah, I'm really excited for this one because you're teaching something that maybe it's not so out there. I think you're going to break this down for us. But to set the tone for this session, What can people expect to get out of this session if they stick around until the end?

[0:55] Yeah, so essentially what they will get out of this session is truly understanding the impact of having a permanent asset inside of the company that is either theirs or a company that they are working for. And when that foundational asset is in place, then the AI agents that we can bolt onto it act as flight wheels with inside the company.

[1:16] Wow, okay. So tell me a little bit how you got into all of that because that's also something I was wondering because I was checking out like in the research, Yes, absolutely. your website and you have quite a lot of different things going for it and maybe not something everyone is talking about.

[1:30] So maybe break this down a little bit and then we'll get into a full screen share walking people through how they can set something like this up. Yes, absolutely. yeah, I've kind of got an interesting journey to doing this today. So I'll give you the Cliff Notes version. So I spent, I used to be an accountant of all things in the past life, spent many years working for Ernst & Young Financial Services in London.

[1:53] And then I moved into the private equity industry for many, many years working. ~ I wasn't a accountant but very much working, running continuous improvement function, product development function, that type of thing. Ran a team of 47 people in there and kind of really honed my skills in kind of process modeling, being able to design architecture and things like that.

[2:17] Not really understanding at the time what I was going to end up doing now. And so anyway, left corporate, what, seven or eight years ago now and I came into this expert industry, as an executive coach because it seemed like a logical thing to do but very quickly discovered that it's the system process automation infrastructure that is my wheelhouse.

[2:41] And was also something that the expert industry seemed to struggle with a bit. So there was a really beautiful kind of dovetail there. And I then moved into kind of what's better known as quiz funnels. I built quiz funnels for years for clients. And then in the last number of years, I've moved much more to working with the absolute elite thought leaders of the world, the big household names, where we build their diagnostic assessments that this becomes this foundational asset that we're going to be talking about today.

[3:11] And then whenever, you know, as AI kind of started to become more mainstream, I've gone down the rabbit hole in a big, big way over the last 12 months or so. So we now have a division in the company that builds AI agents for clients. So we are actively working in this all the time.

[3:31] And obviously the landscape is just changing so rapidly. It is so exciting. There's so many opportunities. And one of the things that I'll be speaking to today is being able to choose really wisely what agents to build first, next and next, because not all agents are created equal and we don't want to just go and build kind of what I call like fancy little toys unless they're solving for the right problem inside of the business.

[3:58] Yeah, and this is really, really cool. And what... maybe you can break down a little bit about your take on AI agents and before we get into the screen share and all that, or maybe you can, maybe you'll do that during the screen share, but there's different types and you know, there's tools that have kind of agents built in now, like even the tools you might already be using, like a Descript for video editing, Mm-hmm.

[4:19] they have agents, but I think what it is that you do is a little bit more on the other end, you know, you're taking people's, you know, their core expertise and what it is that they do and you use that. So I'm interested to kind of hear your take on what it is that you do with the AI agents.

[4:35] Yeah, so I'll be unpacking a lot of the different agents that we have built and why. My view about agents is that it's never really about the tools or the button pushing because essentially that is becoming commoditized and everybody can do that piece. That is not the complicated piece.

[4:52] The most important piece is actually to look at the whole ecosystem of the company, all of the divisions in the company and how each activity relates to the next so that we then, whenever we map out and ideate all of the agents that are possible, we're actually building the correct ones first.

[5:12] So for example, if we are struggling with a lead generation and our market attraction problem, building an agent that solves for client onboarding, why is this really cool? It's not actually solving the right problem for that particular company. So I'm not, I'm not a big fan of kind of cookie cutter, you know, click this button and you're going to have 75 agents, you know, all set up and ready to go. It's bespoke work, it's supposed to be thoughtful and it must be aligned to the specific situation of the company.

[5:45] Yeah, what's your take in terms of how, what kind of tasks can the agent do that? What dollar value would you put on that at this stage we are right now? And then we can get into the screen share. I'm just interested what it is. Like some people are saying maybe $3 to $5 an hour.

[6:02] Some people say more. I mean, what's your take on like kind of what type of employee kind of agent, you know, See take the job from? Yeah, I think whenever we get into presenting today, it's really going to uncover that it's actually an extreme range. So of course, there's the administrative, you know, you know, five dollar in our tasks like the one where I got an email this morning giving me a briefing of here's what I had in your day and an overview about you and your past and your history and how to connect with you on this call, et cetera, et cetera. So those are kind of the $5 an hour agents for sure.

[6:39] But one of the ones that I'll be speaking to later is an agent that we have that listens to every single sales call that our sales team holds, analyzes the patterns, the tonality, how well they did... in relation to converting that sale, where they won the sale, where they lost the sale, what to do better next time, and then feeds that back to the sales team and to me so that we just become a continually optimizing flywheel in relation to, and that's just for sales. So for things like that, where we are actually building our intellectual property, where before I might have had to like watch the entire call and give feedback and think about doing this next time, and here's how to improve that,

[7:20] the agent is now programmed with my intellectual property to then give that feedback to the sales team member so that and then that obviously eliminates that time for me. So that's actually a pretty high value agent. It really moves the needle in terms of the business, terms of revenue generation, improvement in the team.

[7:41] So it's not kind of they're all five dollar agents, there's some of them that are worth thousands of dollars and some of them that are administrative things that are just kind of to-do lists. Yeah, that's amazing. think you set the tone. How about we dive right into sharing and demonstrating how this Yeah, that's good. Let me get my share screen.

[7:58] can work for everyone watching. Essentially what we are going to be unpacking today is that the diagnostic assessment is the foundational asset in any company. And that diagnostic assessment can be something as simple as a quiz funnel at one end of the spectrum, right through to the multi-dimensional diagnostic.

[8:18] Like if you've ever been through Myers-Briggs or disc or strength finder. Okay. And then the agents are the flight wheel. And my, my point of view here is that there are a lot of people that are kind of jumping on the bandwagon, are kind of, sexy, they're shiny, they're fun, this is like the hot topic of the moment.

[8:36] But if we start building agents and just randomly have a load of agents out there operating on their own without any foundational infrastructure, without them actually talking to each other and working together as a team, they're just kind of fancy things sitting on the internet.

[8:54] They're not actually going to improve the business. So as I said, I'm Maeve Ferguson and we engineer proven intellectual property into data generating infrastructure. And in our view, it is the only moat left in this AI world. The client engine is built in. The real asset is the proprietary data that compounds into lasting wealth and legacy. So as I said, this diagnostic assessment is the primary piece.

[9:22] So this is where your intellectual property is codified into permanent infrastructure and every completion of that assessment or quiz funnel is simultaneously a lead qualification event. So it qualifies every single lead coming through. So we know if somebody is a multi six figure client, we know if they're a 50K client, a 10K client, or if you're kind of in the digital product space, might, you know, this is like $47 lead type thing.

[9:50] Okay. So every single lead is qualified by the assessment. And then the transformation delivery happens throughout the assessment where they feel really seen, heard and understood. Okay and the whole idea with these diagnostics and this foundational piece is that it acts as a routing engine.

[10:10] So we're able to get the right message and the right offer in front of the right person at the right time. So we're not trying to sell our multi six-figure thing to the $47 lead and vice versa. So it really, really acts as that foundational piece. And if you miss any one of them, you're going to have lost the sale.

[10:28] So our view is that this diagnostic assessment is the foundation of the business. And the whole idea is that there's actually a bigger layer underneath all of this in what we call the data play. So leads to the byproduct, clients are taken as a given, but the data play is the actual business.

[10:48] Because this data set, it can be monetized six to eight different directions simultaneously. And when I'm talking about data, I'm not talking about kind of selling like zillow wise, like selling people's names and email addresses. It's not that. The diagnostic assessments actually capture hundreds of data points against each lead and when that data is aggregated it becomes a highly valuable asset that many many institutions out there pay handsomely for.

[11:18] So in my view data is the only moat because knowledge is now valueless because you can basically figure out anything you want. Content can be replicated, frameworks so even if you have amazing intellectual property it doesn't take a whole lot to be able to reverse engineer all of that using AI now and essentially have your IP as somebody's consultant inside of their pocket.

[11:43] Courses are completely commoditized and AI can now produce, as we all know, it faster and cheaper than any human. But the proprietary data that these diagnostics build on how a specific population of people actually thinks and behaves and makes decisions, that can't be replicated. So it can only be collected through infrastructure you own.

[12:08] And this is why I'm going to die on a hill for this is the diagnostic assessment is the absolute foundation of any company. So there are many data buyers. I won't get into this because I want to get into the AI piece. But just, know, for those who are interested, there are many data buyers. There are two tiers of data that can be sold.

[12:28] And the whole idea is that every single lead that goes through your diagnostic becomes a data point that has a monetizable value attached to it. Now let's get into how it all connects because we have this if you think through the diagnostic. Okay we have marketing and lead generation agents okay so think top of funnel all of these agents are out there working for you day and daily driving leads into your diagnostic. Now when I say diagnostic as I said we have a quiz funnel we've got a score based diagnostic we've got multi-dimensional diagnostics And all of these agents are actually driving traffic to the diagnostic.

[13:13] The diagnostic then qualifies the leads, brings them through a transformational journey, converts them and captures those data points on every single lead. Then it routes them through to your sales mechanism. So whether you have a by application only or you have sales calls or product sales of digital courses or webinars, that routing engine delivers people to the right message and the right offer.

[13:38] to the right person at the right time. And then as people are going through this, the outputs of those sales conversations, the outputs of the diagnostic are then analyzed by agents as well. So we're gonna unpack all of these in detail in just a moment. I just kinda wanted to give an overview of the flow and the structure.

[13:59] So top of funnel, agents pushing the traffic into the diagnostic. The diagnostic acts as that routing engine, pushes people to the correct sales mechanism based on their investment ability and what they're struggling with, et cetera, et cetera. And then we have agents on the backend that are analyzing all of the inputs, outputs, and then feeding back to the teams to make sure that we are self optimizing for the marketing function, the sales function, operations, et cetera, et cetera. Okay. And then all of the outputs of these are feeding through to what I call intelligence feeds.

[14:34] These are coming through strategic and operations agents. This is where we're compounding the data. We're able to run KPIs. We're able to build pipelines, do forecasting, et cetera, et cetera, because of all of the prior work and all of the data points that have happened inside of this diagnostic. So does that make sense so far?

[14:55] And then what we're going to do now is kind of dive into each of the agents, how they work. And my view here is as I'm going through each agent, have a think about your specific use case. Because there are going to be some depending on the type of creator business you have there are going to be some that are just not relevant for you and some of them where you're like, my goodness that would save me like literally like a hundred hours a week.

[15:20] Okay, so just think through which of these would be good use cases for you and then as I said in the intro make sure that we're not just randomly jumping in and building stuff We want to think through things from this infrastructure this kind of zoom way out to the 30,000 foot view, what is the ontology of the whole entire piece, how do all of our divisions interconnect and work with each other and connect with each other in a manual way currently we'll have funnels and automations and all of those types of things, of course, but how are they all interconnected?

[15:54] Because if you don't do that piece first before you start pushing buttons, what's going to end up happening is you're going to have a load of disparate agents that are off there doing their thing. They're going to be spitting out amazing output. the next agent isn't actually grabbing that output and doing something with it.

[16:12] So we just end up with noise and not actually any improvement in the business. So is there any questions so far or will I keep going? No, this is great so far because you're giving a... mean, you're setting the stage basically for what's to come. I think that's always good. Yeah, okay cool. The way we then do this and obviously there are many many many more divisions in a company you can have finance agents and HR agents and all of the things but what I wanted to do just kind of go through four main divisions. These are agents that we have already built so we know that you know all of what is involved in them, what works, what doesn't work, what to consider as you are building these agents for yourself but

[16:54] you can obviously extrapolate this out across different divisions. well. Finance, HR and so on and so forth. This is just to give the flavor. So marketing and lead in. Interestingly, whenever you look at all of these, this is always that seems to be the highest volume of agents that we build.

[17:12] And this is for our company and for all of our clients as well. And it's interesting that this is where whenever we normally do the ontology of like, here's the overview of the entire company. For most companies, this is where we actually start the build as well, because the bottlenecks which are what we're trying to solve for with these agents the repeating manual activities that take a load of time and don't add much value those are the things that we are trying to optimize the things that we're trying to improve and a lot of them actually fall into marketing lead generation so let's go through these one by one this is one of my favorites so I ~ think similar to you as well I do the newsletter model where once per week...

[17:59] Even though I am way down this AI rabbit hole, I actually write my Substack by hand and I'm not a big believer in kind of spitting out AI generated garbage because the world is full enough of that. So we write that by hand, or I write that by hand every single week. But I don't have time to sit there and like repurpose it or even like run it through Claude to say like, right, turn this into a LinkedIn post for me.

[18:22] What we have instead is an agent that repurposes all of that content across all of the different platforms and the whole idea here is and this is like an important point to note that we're not just bumping this into like any random agent and say like repurpose this content the agent is pre-programmed with years of my intellectual property it is pre-programmed with my brand voice that we have worked incredibly hard to refine to optimize to improve all of those things every single week the IP by gets updated with every single training that I have done.

[19:02] I do about five podcast interviews per week, any coaching sessions that I had, any consulting sessions that I've had with clients, any VIP days and so on and so forth. We have a process to grab all of that IP in every single time I open my mouth, every single time I type something, it goes into that IP bank that then informs the content repurposing agent so that it literally comes out as though I wrote it and it's not allowed to use any AI generated words it's only allowed to use the words that I actually wrote inside of my newsletter.

[19:40] And that allows us then to be omnipresent across all of the platforms. And essentially all I have to do in any given week is write my Substack, pass the team the Google Doc, they run it through the agent and everything else is kind of done and dusted. Okay. So really, really, really useful in terms of the number of hours saved, both from a kind of a VA perspective, but also from the, the, the creator themselves perspective, saves a colossal amount of hours.

[20:09] Okay. Social media intelligence. This one's pretty cool. So this is where obviously all of this content that you have just created and posted and put out there, some of it's gonna bomb, some of it's gonna land, etc. So what we want to do is we don't want to just do things for doing them sake.

[20:26] We don't want to just post across all of the platforms and not actually learn from what our market is telling us. So my highest value is continuous improvement. I always want to know what can we do better even if it is the tiniest, most incremental little thing, if we can improve that even 0.1%, it's still better than we were yesterday.

[20:48] So for social media intelligence, what this agent does is it then goes and reviews automatically all of the content that was posted prior week, lets us know the reach, the engagement, the likes, the comments, the interaction or lack thereof and then summarizes like here's what landed, here's what didn't land, here's what to do better next time and then feeds that back to the marketing team and myself so that whenever I am writing next week's Substack we can do a better job.

[21:17] So it just becomes a self optimizing flight while. So I absolutely love that one. Podcast and YouTube script agents. for our podcast, sorry, for our YouTube channel, I'm not a massive fan of completely scripting it and you're like reading like a robot off the screen. But I do like to have like, you know, an outline of like here's where to take the conversation.

[21:39] So this agent essentially we just dump in a topic and again pre-programmed on the My intellectual property my point of view my model of reality This then pushes out the outline or the script if you prefer the outline of here's what to talk about in your video so that it captivates and converts.

[22:01] Okay, so you can build these in any way that you want the common thread that I want you as the viewer to pull on and to hear is that anybody can spit out a YouTube script having it built with the correct IP in mind having it built as a machine that knows how YouTube works, knows how podcasts work, knows the mechanism by which you're sharing the content so that it actually converts is the most important piece because pushing the buttons is the easy bit.

[22:32] Email campaign copywriter. So I'm not going to go through all of these in like great, great detail. Otherwise we'd be talking for five hours. But again, all of these ideas just have a think for your business. Where I see a lot of people who are, yes, I'm using AI. They're just having a conversation with Claude or ChatGPT or whatever platform they use.

[22:50] Having agents that can actually take the piece that we want to do. We put an input in and it does all of that forward and back and act as just sparring partner inside of the agent. So then all we have to do is then just collect the output. It's a very, very different thing than just kind of manually going forward and back using AI. Okay. One of my absolute favorites is the content calendar planner. It's like right here's, and especially if you're in an evergreen business model, works for launches as well, you literally just put in here's, here's what's coming up. Here's what I want to be focusing on, on the next six months. So for us inside of our company, for example, we have have the diagnostic assessments that we do done for you.

[23:32] We run VIP days. We have smart scale, which is our agentic AI done for you service. So I want to be able to get into a good cadence of sharing different offers at different times of the year based on what our goals are as a company. And that content calendar planner allows you to do that.

[23:49] So you're able just to give it instructions and then at the backend out pops your beautiful content planner. Headline hook generator. Content ROI attribution... So this one is pretty cool. So again, it kind of ties in with this social media intelligence, but this is like which pieces of content, which Facebook ads, which copy, which creative actually resulted in sales.

[24:11] And this is the kind of the powerhouse of the diagnostic assessment where we're able to, we're able to like use this routing engine. We know exactly what piece of copy, what ad, what campaign, what ad set they came through. It came through the diagnostic, they then bought or they didn't buy and if they bought, they bought at different levels.

[24:31] And what we're able to do then is know which pieces of content actually turned into an ROI inside of the business. So super, super cool, really, really valuable. And again, all of these things talk to each other so that they aren't just disparate agents of kind of working randomly by themselves.

[24:50] Brand voice monitor, cool. It's not going to be for everybody. But for for those who have really really high profile where the know, the online streets are kind of talking about them this monitor keeps a check and balance of any mentions of you as the creator online and flags to your attention anything that's like this could be damaging to our reputation or and the good things that are happening in out there as well just so that you have a finger on the pulse.

[25:19] You're not waiting for you know, somebody to let you know this somebody has said something bad, you're just informed on a daily or weekly or monthly basis or whatever cadence you choose to understand how you're being seen in the market. So Hot Leads Alert System, I absolutely love this one.

[25:36] So if you think about leads coming into any company, are leads that get forgotten about, there are leads that just kind of vanish into thin air, there are, and not all leads are created equal. So we have some people who would be better being served through watching your YouTube channel and some people who want to work with you at a multi multi six figure level if not more.

[25:58] Okay so there's a wide spectrum of what can happen in there and what we don't want to do especially if you have a sales team ~ that type of model is you don't want your sales team focusing on leads that don't have 20 bucks to spend. Okay, so the whole idea with the Hot Leads Alert system is that it knows as somebody's going through the diagnostic, it knows based on past patterns of behavior, past investment markers and wealth markers that we build into these diagnostics, who's actually gonna be the person that can afford to work with us, is gonna actually jump into our program.

[26:37] Okay, and again because we have so many data points on every single lead that we are Analyzing in the long term. We're aggregating that data. We're understanding the patterns of behavior in every single market that we are in this hot leads alert system knows how to differentiate a hot lead from a luke warm or a cold late and then what that allows us to do is to send a slack notification to the sales team to say so and so it's just opted in...

[27:06] This is a VIP kind of hotly tier one, go look after them. And the people who are coming through that probably not gonna invest in us anyway, we're able to allow the automations to handle them and just take them through the systems. So that's a really, really valuable one. This one here is pretty cool.

[27:24] Kind of weird, but cool. So basically what this one does is it knows, even if somebody hasn't opted in, ~ it knows when somebody hits our website, when somebody checks out our application page et cetera. What it does is it gives us their name, their email address, their phone number, their LinkedIn profile, all of these things.

[27:44] And then it goes and auto follows them or connects with them on LinkedIn. So it basically acts as a connector for you. You can set it up to send the messages and all of those types of things. I personally prefer the personal interaction. So we stop it there, but it is possible to then start sending them DMs and all of that kind of stuff as well.

[28:06] But it really gives you really good insight into who is coming into your world. It's always hilarious whenever I have a podcast episode air and you can see this flood of viewers coming in from a very specific vertical and this agent basically tells me that the episode is aired before the host of the show has even let us know that it has been aired.

[28:28] Okay so super super cool and it allows you then to engage with people who have been checking you out. Speaking Engagement Scout... really really cool. We actually had to turn this one off for a while because it was providing too much value and this Speaking Engagement Scout is one where it is given very specific instructions about the types of people that we want to work with. Okay, so who do we want to have in the audience?

[28:55] For me, it's like really elite thought leaders. Okay, so think through who you're like as a viewer of this, who is your person that you want to get in front of and then the speaking engagement scout goes out there and finds speaking engagements both virtual and in person both free and paid.

[29:15] And then essentially it does an analysis based on pre-programmed instructions, grades all of the speaking opportunities from A through F so that the team then know, right, let's apply for these three speaking opportunities. But these ones here we can just completely ignore. And it also looks at my calendar and it knows if I'm in New York or Ireland or wherever I am, and it is able to recommend speaking opportunities that come up as available whenever the agent knows that I'm already going to be in New York that week. Okay, so super, super cool. But we actually had to turn it off because it was for a while because there were so many opportunities came through, even at the kind of the ABC grade that even if I did keynotes like

[29:54] you 25 hours a day, there just literally would not be there is just still too many opportunities. So little health warning with that one. Application Review Agent... So I love this one because it kills a lot of time. and manual effort in the sales team. Okay so application review, is this a fit person?

[30:21] Is it not a good fit person? Which one should be led through to the kind of book a call with us? Which one should we pass and just kind of send them some freebie just to keep them going? So this agent helps take all of those inputs. Again, they've been through the diagnostic, they apply to work with you and then this agent reviews the quality of the application and then lets us know whether to go ahead or not.

[30:44] Lead tier classification, I talked about this a little bit earlier already, where every single lead is classified into platinum, diamond, gold, silver, bronze, or unqualified. And we're able to then just send them on their merry way, depending on the caliber of lead that they are.

[31:01] Lead nurture manager, diagnostic intelligence. This one is where we're able, the agent with the diagnostics is able to monitor completion rates, drop off points, It analyzes the quality of the data, tells us insights into our audience, that then informs all of the other agents that are working in the marketing division, top of funnel.

[31:23] Okay and then LinkedIn Connection Builder, this one is super cool and useful to pretty much everybody. So essentially this is instructed with... Here are the here's my ideal client here. So for me, it would be bestselling author keynote speaker, TEDx, all of that type of thing. And it goes out actively every single day and builds my audience for me on LinkedIn, connects with the people, sends them an initial message, likes their content, etc. And then when we have sent that initial message, then we take over then and have real person to person conversations.

[31:58] So that connection builder, again, hast just replaced an entire VA that is now just running every single day, day in, day out. And it allows us to build that network and then that turns in obviously on the backend into clients. So those are a few ideas for marketing agents. And there are many, many, many, many more. This is just kind of a, a kind of a high level overview of what is possible.

[32:27] So hopefully the wheels are turning about like, ~ that would be so cool. That would work for me. That one, not so much, but like these six, for example, would be the first agents that I would build because they're solving for a bottleneck or a problem or a struggle that I am experiencing.

[32:44] Okay, so that's the marketing angle. Let's move on to sales a little bit. So I talked about earlier. Yes, let's unpack a few things here. I mean, this is really good because you said you're actually handwriting your newsletter. I just kind of caught up on that because Ryan Leveque, Yeah.

[32:59] he's been on my summits and workshops in the past and he says he does Yes. the same for his newsletter every week. But then as you do, he takes that newsletter that he handwrites and Yeah. he turns that into pieces of content for his YouTube and other platforms, Exactly. right? And even I think he has newsletter to his main list and he has a Substack and he has different ones Good.

[33:21] and he just, mean, I'm sure it's maybe his team that is doing some of that with the help of AI, but maybe we can get into as well, kind of maybe we won't have time to get into everything in detail, how to set it up, but like what tools would you use, for example, to do certain things?

[33:37] think the audience would be super interested to know kind of what you would recommend, at least there are many ways to do things with AI, Yes. but kind of your recommended stack for doing certain things, especially the once in content creation. Yes, absolutely. So we are ~ focusing on written pieces for now and I use a range of different platforms ~ but ours is slightly different because I'm always building agents from the viewpoint of usability for our clients because we actually build agents for clients. Okay, so I did a lot of work for about nine months building everything on Manus. And I actually ended up, and again, I love Manus, I use it all the time, it's incredible. But I actually moved to Lindy for the done

[34:30] for you agents component of our business because it is easier to get those agents across to client accounts and get everything set up and for it to be usable and understandable for them. So we use Lindy quite a lot. Then what we do, and I learned this the hard way, I was telling Mike the other night that I had like a lovely multiple thousand dollar lesson learned in my early days of playing with this because I was designing and building all of these agents on Lindy and I like mean munching through tens and tens and tens of thousands of credits.

[35:10] So I learned the hard way ~ for about that. So what we now actually do is I use Claude, both Cowork and the browser version and we design, ideate, design, build go forward and back in Claude to design the perfect agent and then we actually take that to Lindy then to build it out.

[35:35] Okay, so it's super super simple structure. Lindy is very kind of self-explanatory platform if you're building enough the viewer is building this out for themselves So essentially what we are doing super simple kind of stack I'm writing the newsletter on a Google Doc. Literally, I used to like put it through my fire it over and Slack to the team into the content channel and then they have an SOP that they follow to put it into the agent.

[36:00] Okay, so put it into the agent and then essentially what that is doing is creating each of the different pieces of content for all of the different platforms based on LinkedIn rules, Instagram rules, email etc etc. And the next stage on from that note, I haven't gone live with this yet even though it is possible, is where this is all auto posted, auto scheduled, everything. I still have a human being checkpoint in my process even though you can go and cross-purpose this across all of the different platforms. It's physically possible now whereas a few months It wasn't it is physically possible now.

[36:37] I still have that check and balance just to double check the quality of the output Make sure that it is kind of like we're happy with it So that is the route that we have chosen to go and say we're maybe a couple of months off. Been able to automate like we use HighLevel so like being able to do all your scheduling through that pulling from an asset bank of you your latest photo shoot and all of that type of thing.

[37:01] So that is... we essentially use Lindy to build these agents and then another really good lesson learned that we had was we used to build all of the instructions if you like say for example for this one all of the instructions we used to build them into the agent itself So the brand voice and here's my like all of my IP, etc, etc. And what we soon discovered then is like as you try to roll this out to other people, there's you're basically rebuilding the wheel every single time. So now we just have kind of documents on a Google Drive with each of the inbound assets that the agent needs to pull from that then those can just be updated any day. And then the agent, the next time it runs, is going to grab for the from the correct document.

[37:48] So having that structure set up All If you're using Lindy or it doesn't really matter... n8n or Make, or Gumloop or whatever, in my view that the tools are the least important piece. But it's what everybody thinks is the most important pieces that everybody wants to talk about is like what tool are you using, Yeah, I mean, I you know?

[38:07] do think you can use also Claude Code, you can use a lot of different things, Yes. but I think like Claude Cowork is not as autonomous as Lindy. Lindy can just go out there and do the work with Claude Cowork, you kind of have to be a little bit there, you know, and it does the work, Yeah, exactly. Yeah. Yeah, but you have to like approve stuff and things like this.

[38:27] So that's kind of how Claude Cowork works. exactly. Yeah. And I think Claude Code is a little bit on another level and then obviously with OpenClaw and things like that, it can just kind of work for you. But I think what would be interesting during this session also talk a little bit about the cost to run certain things.

[38:41] You mentioned a bunch of different agents here. Yes. Maybe if you can and it's not confidential, maybe going into a tool kind of showing some examples of kind of what an agent can look like, especially maybe the repurposer you mentioned for your newsletter. That could be exciting for a lot of people watching.

[38:59] Yeah, so the cost is an interesting piece, especially if you are building for others. But it has to be, even if you're just a user of an agent, you've got to understand that it's not free. So there's some people who like, what do you mean? I didn't realize this agent was going to cost money.

[39:20] And it's like, well, how did we think it was going to run? So obviously they use credits. But I think it's important that whenever we are, again, back to that, that high level overview that I talked about earlier and deciding what we're gonna build first, next and next, we want to be really mindful of, like right, if this thing is gonna cost like a hundred bucks to run, like it's probably not gonna cost that, but like let's say it's gonna cost a hundred bucks to run, you're gonna run it every single day, is that actually in budget for your business?

[39:51] So say for me for example, where's the one, the brand monitor, okay? Very little value in running this thing every single day. Okay, because you know, I'm not Mel Robbins... okay, there's not like people commenting on myself out there and you know sharing articles about me every single day So for us, this is like a once a month activity.

[40:12] So you want to be mindful of the cadence that you run your agents at again depending on the type of business you have the profile you have the volume at certain kind of intersections in your business that you have you want to be mindful of all of that and as you're building these agents knowing so one thing that we're starting to do as a team because obviously we're building these for clients and rolling these out for clients we're actually able to tell them then like right this is how much this agent is roughly going to cost so say for example our LinkedIn connection builder it's about five dollars a day okay so we're able to actually then they're able to ~ kind of compute how much that's going to cost them over the course of the month and they

[40:56] say yes we want this one or no we don't So really kind of understanding the cost per run comes from two things. So every single person in the world is like currently kind of building the plane as they're flying the plane as they're building it. Okay, so we're all this is all brand new for everybody.

[41:15] And for us, it's just been very much build it, run it, and then go back and look retrospectively to see almost like how many credits did it gobble up. And then obviously you can optimize these things as well. So you don't need to be using Opus 4.6 if it's just a read and understand function.

[41:34] You wanna make sure that you're pulling on the right model depending on the complexity of the task that the agent is being asked to do. So that's a really, really important piece from being able to decide what you want to build first, next and next. Do you want to go to Lindy or will I just kind of run through a few I think we run through this first with your of these other examples?

[42:04] agents because you mentioned the sales. I think it could be good to just go through this and then we can go into the tools and show people inside. Yeah. I think that's always good. Although the tools can change, think the fundamentals that you are going through is really, really important.

[42:19] Yeah, okay. So sales and conversion intelligence. Okay, so this one here is definitely my favorite. ~ Reason being is that we have programmed the agent on kind of global best practice sales frameworks so that the agent knows ~ whether somebody has done a good job in the sales conversation.

[42:42] So for example, if they were going through the band framework, we have this built into the agent and it's like, okay, you did this really well, but you forgot to ask, you know, what's their budget? Okay, so like a really basic one-on-one thing in sales, we didn't actually get the budget and the agent is reviewing the transcript of the sales call automatically as soon as it ends, that triggers the agent automated in the background.

[43:07] That agent is pre-programmed with instructions to listen to this transcript, compare it against best best practice framework A, B and C, give feedback to the sales team member on what they did really well, where they have room for improvement, if they won the sale, how they won the sale, if they lost the sale, how they lost the sale, if we have a two call closed process because our work is kind of requires that to go and build a bespoke proposal.

[43:37] So we then have it built into our agents. It's like it gives like here's the likelihood of this converting at the next stage of the process. And it is eerily accurate whenever you look at the actual close rates across the business and what the agent said was the likelihood of this converting.

[43:54] It's pretty much exactly where you would expect it to be. So this agent is like it's doing objection handling, it's spotting patterns of behavior, closing effectiveness, how much the person the sales rep is talking versus the lead talking all of that type of stuff but interestingly this is...

[44:14] It really kind of us laugh internally in the team. It's an example of agentic AI improving over time. So whenever we initially built this agent and it was running, it had zero capability of reading the tone of the room. Okay... so I'm not involved in sales in our company at all. But as this agent was rolling out, I was super curious. I was like opening up the output documents that the agent created, which was the report about the performance of the sales call.

[44:42] And I was reading that. I was like, this person sounds so horrible. Like I don't want to work with them. You know, and I was almost like, this kept coming through with where it made the lead sound like a really difficult human being. So I was kind of going back to the sales advisor and saying like, oh gosh, like they sounded like a nightmare.

[45:01] Like, hope you got through that okay. And they were like, what do you mean? Like they were really lovely. So I got curious. I then watched the replay of the sales call. The person was perfectly lovely, but what the agent actually was the ability to read it, to have the tonality and know the of the chit chat that was happening and the mood in the room.

[45:23] So now we're optimizing the agents to be able to understand that level of nuance as well. So really amazes me how even the things that we built last month, they can already be optimized now. It's just, it's going so fast and things are just getting so much better and better and better that, you know, we'll look back at stuff we did a few months later almost laugh at ourselves that that's where the world was back then.

[45:48] Okay, so sales call intelligence super super cool happens automatically after every single sales call. Proposal creator so we do a lot of these artifacts as you can see so this is just built on cloud code this entire page and again for our proposals what this agent does is it grabs the transcript of the conversation between the sales team member and the lead and creates It's a proposal which we send out a bespoke proposal for every single person because the work that we do is bespoke.

[46:20] It's not a digital course or a cookie cutter coaching program. We have, you we work privately with people. So it is a it is a bespoke proposal. And you remember back in the day when you had to spend hours and hours and hours crafting these proposals. This is now a an automated process, I would say to about 85, 90 percent. It still requires a little bit of refinement, okay, so it's not, you know, not going to blow smoke and say like, ~ this spits out a perfect output every time.

[46:50] You still got to run your eye over it and make sure that it 100 % aligns with what was agreed inside of the initial sales conversation. But it collapses, just gets rid of all of that time. And back to your initial question at the top of the conversation, this is where the, this is like super, super high value.

[47:10] So if you think of the cost of somebody having to spend like seven hours like you know laboring over a manual proposal this eliminates like the vast vast majority of that work okay so it's just run your eye over it we have an internal approval process and off it goes the next day so super super cool conversion analyst so this one is I start it identifies bottlenecks across all the funnel and the touch points okay so where lead and this is because we have the diagnostic okay so as I said before, the diagnostic assessment is the foundational asset in the business. Everything that happens, all of the agents that drive leads into it, are so that we can get them through the diagnostic and then the diagnostic acts

[48:00] as that routing engine, but people fall off. So what the agent does is it's analyzing like question 17, you're losing, you know, 19 % of your market as they get to that point. So then we know exactly what to and do and fix and optimize instead of scratching our heads and wondering what's broken in our funnel, we're able to actually know exactly where people are falling off.

[48:23] We build these KPI dashboards and reporting, cetera, et cetera, using agentic AI that allows them to optimize every single touch point of the funnel. Like if you're kind of running organically only and, know, know, hundred leads going through your thing a week, that's one story. We have clients who are like running a million dollars a month in ads through these diagnostics.

[48:49] So we don't exactly want to like not know what's going on. So the value of all these agents is obviously exponentially higher when there is significant ad spend on the front end. High-tech at buyer finder. This one is a weekly report that an agent runs that identifies pattern recognition across the pipeline. Okay. So it touched on this a little bit earlier but this is the one that identifies behavioural signatures of high ticket buyers so that we can find more and more of them and less and less of the people that just want free stuff.

[49:27] Okay, so really really cool, generates a report, it to the sales team over on Slack so that they know exactly who to follow up with etc etc. Follow-up sequence managers, pricing calculators, client satisfaction monitors, There's quite a cool one as well where this list this agent listens to our client conversations It's very similar to the sales call intelligence agent, but this is actually listening to what is being unsaid okay?

[50:00] So what what are people not saying? What are the the things that are happening between the lines that the naked eye couldn't see and the whole idea of this one ~ is to identify potential churn. I actually got this idea whenever I was ~ reviewing a membership site, okay, so where churn is a big big thing, completely different model. So we were like high tickets, like bespoke private work membership is a complete opposite end of the scale but it was the same principle, okay. So that's what I'd like you to take away.

[50:32] It's like don't say that's not relevant for me it's like right there this agent was analyzing potential churn in a membership what is the relevant use case of that for my company and for other companies and it's so basically it listens to see is the client truly happy are they really really loving it is there anything that they want to say but they're not quite saying and because the agent is programmed to understand psychology it is then telling us by the way you just had this call with you know your client today for example and this is going well and this is going well but you might just want to check in with her and see how she's feeling about X.

[51:11] So then we're able to take the four road of that, check in with the client she's like oh my goodness amazing I was wondering about that and we have solved for a problem that didn't never actually got to raise its head okay so we're able to solve things ahead of time and then testimonial opportunity finder one is quite cool as well so same principle it listens to the conversation and it knows what, basically we build testimonial flywheels.

[51:38] So it's like, what is a one-liner that a client said, or, you know, I was on with a client the other day and they were like, yes, they sent a 50K client. Okay, cool. That is an example of what this agent grabs and then says, can you go back to client X and get a formal testimonial about them sending this 50K client? So what it does, again, it's so easy to get so busy and you're so in the weeds and so busy doing, that you know you need to do these types of things but they never actually get done.

[52:09] What the agent does is that it makes sure that actually gets done so then your team can go and action it. So those are some of the sales agents and some of the strategic agents that we have in place as well. So these are kind of very, what I want to say, they're very kind of output focused, they're giving us stuff that is super valuable, super, super helpful in terms of getting more leads, and making more sales. The strategic pieces are more kind of like what I would view as the CEO level activity.

[52:42] So the morning briefing agent, it is pre-programmed to, it knows exactly what me or whatever client we're building this for needs to know, wants to know different topics, different angles, world events, et cetera. I am the type of person I just point blank refuse to watch the news, I don't watch TV, I don't want any negative anything anywhere near me, but I also don't want to walk into your room and you know not know something massive that has kicked off in the world.

[53:10] So this briefing agent goes off, scrapes everything that I've told it that I want to know and need to know, condenses it down into super bite-sized pieces, brings me back and tells me exactly what I need to know so that I can keep my blinkers on, stay in my energy, do thing without getting sucked into the vortex of what seems to be going on in the outside world. Okay?

[53:33] Really really cool. Again, it's not going to be needed for everybody. It's just an example of a use case that for me personally, it solves for that problem that I just don't ever want to watch the news. Okay? Conquest Intelligences... this is I'm pretty sure my absolute favorite one.

[53:51] So we have a running list of all of the dream clients across the world that we want to work with. We're talking the biggest, biggest thought leaders. We're working with some of them already. We want to work with these incredible people. And what this Conquest Intelligence agent does every single day is it is scouting all of their social profiles.

[54:13] So their YouTube, their podcast, their LinkedIn, etc. Substack, etc, etc. And this Conquest Intelligence agent then scrapes all of these to see where against obviously the list of people that we have given the So these are the hundred people that I want to work with It goes and finds where they have shared content where they have shared You know or they've published a book or they've got a new YouTube video or whatever it may be and then it sends us the link Okay, it auto likes it Okay, I'm not a huge fan of the bot comments under the thing because the person knows that it's about commenting But what it does is it actually shares with us the link to the article so we can go in and leave a thoughtful

[54:57] comment. The agent gives us an example of like here's an angle that you can use in your comment. I just prefer to go and then hand write it so it's thoughtful. But this is working across all of these dream client lists. It's keeping us abreast of every single thing that is going on with the people that we want to work with so that it's putting it putting it front and center in front of us every single morning so that we're not sitting there going yeah I must to go and like check so-and-so's profile and see what they're up to these days, it never gets done. So the agent actually drives the activity.

[55:34] Then we've got SOPs and team etc doing certain activities on these pieces. But this Conquest intelligence agent just provides an insane amount of value to the team. KPI dashboard manager. So I absolutely adore this one. Now this one is semi-autonomous and we talked a little bit earlier about some of the agents that are completely automated.

[55:57] You mentioned Claude Cowork, Lindy and things like that. So not all, in my view, not all agents should be completely autonomous and not all activities should become agents either. Okay. So we want to be really mindful about what we are making completely autonomous, what we are keeping semi-autonomous, i.e. where the agent is doing 80 20 of the work and then a human being is doing the final pieces and this KPI dashboard manager is an example of that where we have semi-autonomous processes in place that we and we do this for a range of different clients they just have us on monthly retainer but we basically build essentially Claude artifacts of all of the KPIs inside of their company so we're maybe extracting it depends on the complexity

[56:45] of the company we are extracting data points from like maybe 15 different platforms and when nobody has an overview of all of those numbers across all of the platforms, it just ends up a big hot mess. What we do is actually we have a very, very structured KPI reporting system where we extract data point A, data point B, data point C. It then goes into this kind of process where then at the back end we're able to spit out KPI.

[57:17] APIs for the for the entire company and you're able to dig into your your ROAS etc etc. You're able to even like we analyze down to the level of here's the copywriter who had the greatest ROAS in the team. Here's the piece of creative that worked the West. We are able to analyze any piece of data in the company and then present it in a beautifully... an aesthetically pleasing format so that the user of the dashboard.

[57:47] They can see all of their key numbers. They know exactly what ads are performing, what ads are not, what to turn off, what to scale, et cetera, based on these KPI dashboards. So that's a really beautiful one, but it is semi-autonomous because there are so many inbound data points that there's a human required just to start the process.

[58:07] Competitor watch is very, simple. It does what it says in the tin. It's basically watches. we give the agent instructions of the different competitors that we want to keep an eye on and what they're up to. And then it sends us a report every week to say, here's what's going on in your market, this competitor has made this move, you might want to be mindful of this.

[58:33] And we've also set it up if there's anything kind of urgent, it doesn't wait until the following Monday to tell us, it kind of flags it on the day. So you can build these to be either running on a calendar schedule or to let us know as soon as there's of significance that happens.

[58:49] Revenue forecasting agents, ad performance optimizer, again, one of my favorites where we are so many people who are running paid traffic. It is like spray and pray, whack them up and hope for the best. What this actually does is analyze down to the nth degree of the performance of the copy, even tracking like who wrote the copy, the performance of the creative.

[59:14] Obviously then we're running them through these diagnostics so we have all of those data points as to then who bought what at the back end and then it feeds back to the marketing team to say this copy is working really well, this copy is absolutely booming this particular copy had a way higher start rate than this copy and so on and so forth so then the the media buyers just get this self-fulfilling kind of flywheel of information getting passed back to them as to here's what's working here's what's not so really really really powerful. Quarterly review generator again is kind of tied in with the KPI dashboard but it lets us know on a quarterly basis how everything is going.

[59:53] And then strategic recommendations. So this is really really powerful in that we have programmed it to act as a business strategist but where the value comes in is that we are layering in the data that we have off the back of the diagnostic, the data that we have off the back of the sales mechanisms so that it's able Yes.

[1:00:14] to tell us where to go as a company. So very, very kind of high end, super valuable, potentially not for people viewing right now, but I'm just kind of giving an overview of different things that we are using. And then operations kind of daily pipeline prioritizer. Again, we want to focus on the hot leads.

[1:00:35] We want to focus on those platinum leads that are actually going to pay us instead of spending attention on. qualified leads. This one gets everything into the right order. Task assignment optimizer as well. So again get and think like we all have 1500 things to do. What are the most important things that have to get done today so that we are actually working on the most important activities.

[1:00:57] Deadline monitors and alerts. this is depends if the creators who are viewing, if you're viewing this and you have a lot of deadlines, this is like super useful where it kind of scans what's going on in your project management platform of choice. And it's like able to say to you like, you know, heads up...

[1:01:16] Here's what your deadlines are today, but here's what's coming at you next Thursday and it just keeps you abreast and being able to run really really really strong client projects and then team capacity analyzer. So this one's good for us because we have an agency model and anybody who's running an agency will know that team capacity becomes an important piece whenever you're completely booked out. So you want to kind of, this agent is like really understands each person's capacity, what they're able to work on, what their skill sets are, what projects are in the pipeline, what projects are ongoing, which ones are about to go live, etc. And it's able to fly to us where we are like, okay, You know, ~ Genesis is like literally about to cap out.

[1:02:00] And it just gives you that foresight before agency staff then, you know, burn out and quit basically. So pretty cool. Those are the different agents that I kind of wanted to share with you today. and let me just pull up I'll just stop sharing for a second and come back to you Yeah, yeah, this was, this is amazing.

[1:02:22] I'm sure people got so many ideas. I mean, I've never seen someone share so many different types of agents and like one sitting like this. That's awesome. Yeah, but it's just ~ scraping the surface. There are ~ a million things we could build. I feel very strongly just because you can build an agent doesn't mean you should.

[1:02:43] And it's like, are we actually solving for the right problem? I think that is the most important piece. So I'll just pull up, for example, our marketing agent. Let me see. Yeah, if you can dive into a few to kind of show on the inside what it can look like and maybe how you decide.

[1:03:02] You say you use Lindy for most of the stuff now, but do you use any n8n setups and stuff like that as well? Yeah. I mean, that's been popular. I mean, Yeah, same time. I actually something I saw from Sabrina Ramonov, she's a speaker on the summit as well. Yes. And she is the founder of a tool called Blotato.

[1:03:22] I'm not sure if you have seen that, but that can automate a lot of posting to different platforms and she blew up on social media using Yeah. her own tool basically and of course posting good content but that's kind of what it's for to amplify kind of what the content you create already and then you can connect that to Claude Code or whatever agent platforms you are using and with n8n and stuff like that so I thought that was pretty cool.

[1:03:44] Yes, fabulous. So yeah, so what I was saying, sorry, is like the, what amazes me and what I love about AI is that, you can put like a hundred of us in the room who are doing this agentic AI rabbit hole. We all have different points of view. I this is why the summit is so valuable.

[1:04:00] We all have different points of view. We're all playing with different toys. We all have different experiences, different insights, different things that we have built. And the beauty of it is that that level of awareness, like even that platform name, like that's a writer downer.

[1:04:16] Okay. Cause then you can go investigate that. The example that I gave earlier about the sales call analyzer, where I just wasn't getting the tone of the room. I was at an agentic AI event in Boston there a couple of weeks ago and a guy, was random, I don't know why we ended up chatting about this particular agent and he was able to give me the name of a platform that I didn't know existed that was able to assess for this tonality.

[1:04:42] So now we're able to, we're all actually building that into that agent as we speak. So it's like all of these conversations, every single speaker on the summit is going to drop like a platform or a or a one-liner. Or a different angle or a different viewpoint as to how to use agentic AI that like it's the beauty the beauty of the summit so let me see like let's go content repurposing machine for example okay so anybody that's using linda you'll be very familiar with this okay so this was ~ run this morning at 7.27 okay so you literally ~ have and i'm sure everybody's aware of how this works i won't get into like how to push the buttons ~ but basically it's just like how to get started and then you just we just dump

[1:05:29] in the newsletter that I have written and then the flow editor is where you'll see how this all works okay so the begin the chat is the piece that we just saw then it essentially has to go um let me just close of that. So it has to go to gather all the inputs. Okay. So it's like, here's, this is like the instructions that we put in.

[1:05:51] So it asks us basically what week are we in a notation? We have five different plays that we have in terms of our content. So it basically the agent asks you, are you in week one, two, three, four or five? Okay. So for this week, for example, the newsletter was all about agentic AI, which is our offer that is called smart scale, which is week number three. Okay.

[1:06:12] So we say it's smart scale week three. What's your Instagram keyword? So where you, you know, DM the keyword off the back of the post. Okay, so I just said the word was agent. Super simple. And then this week's URL was you're putting in the link to the application to come and work with us in smart scale.

[1:06:28] Okay, so you literally just put in those inputs done and then the agent has these instructions to follow. Okay, then this is the important piece. So this is where I used to do way more work than we needed to, where we used to have all, so these used to be like 19, 20 steps long, okay, used to bug out and things like that.

[1:06:53] So we are instead of having all of these instructions built into the agents, which is what we used to do. We now just send the agent off to go and read the strategy doc. Okay. That has all of my, my brand voice, my IP that we update every single week that I talked about at the start of the call.

[1:07:13] It goes off and reads the strategy doc. Meaning that then whenever we copy and paste this agent and roll it out to a client, all we have to do is build same doc for them. And then connect the agent to their doc instead of having to rebuild the wheel every single time in here. Makes it super super simple. Then this is when it moves in and this is all completely automated. This is where it follows the instructions of the cadence of the content I want to produce.

[1:07:42] Okay so this is going to be completely different for everybody. My Substack goes out on a Sunday. I want an email to send people to my Substack. I send three emails a week and I want to drive them to the call to action for the week's main offer which in this week's example is smart scale or agentic AI done for you service.

[1:08:00] Then daily social posts okay so we are doing ~ Instagram, Facebook, LinkedIn, Substack Notes, all of the usual suspects etc okay and then so these are basically the instructions that the agent is following so that the content gets repurposed in the way that we want it to be done.

[1:08:20] It doesn't sound like some AI ChatGPT generated garbage, it is literally built on our brand voice. Okay, so no em dashes ever, British English spelling, because I live in Ireland. Only text from the newsletter, don't invent anything, don't hallucinate, et cetera, et cetera. So we're putting these instructions in, Thanks.

[1:08:39] okay? And then super, super simple. We have so we actually built this QA agent in Off the back of this because it used to go from here straight out Now we have a QA agent that its only job is to check the quality and the caliber of the work created okay, and then when it is done, it just sends a slack notification off to our content channel in slack with the link to here's all the blurb and then my team ~ who do my posting etc pick it up from there?

[1:09:11] Okay, so... This absolutely can go to the next stage where it just auto posts as well. It's just I still prefer to have a human being just run their eye over it and make sure that it's kind of where we want things to be from kind of my brand perspective. Okay. Just a question on that one. Do you have Yeah.

[1:09:34] some examples of the platform's content? For example, on LinkedIn, it's different than X, Yeah. Yeah. for example. I don't know which platforms here are the most active, let's say, do you have like... Yeah, like that. Yeah. it's searching automatically or have you uploaded content that you cannot Yes.

[1:09:49] get access to? Yeah, so that's what happens inside the strategy document. So the strategy document is essentially here's all of me as IP, here's me as brand voice, here's how you may have speaks. OK. Or whatever, whatever the client is. And then it's just like, here's what's best practice on Substack notes here, it best practice on Instagram, Facebook, LinkedIn, et cetera.

[1:10:09] So it is following essentially the rules. It's not, for example, putting a clickable link into the body copy for a Facebook post. It's not putting, you know, it is putting hashtags onto an Instagram post, but not Facebook. Okay. So all of those rules are built in to this strategy doc.

[1:10:29] Yeah, that's great. also, that's kind of like the skills that are now very popular in Claude. It's kind of similar. Yeah, skills are amazing. You have the skills, and then you can also have plugins now in Claude Cowork, and things like that. So it's quite similar how you need Yeah.

[1:10:45] to essentially give whatever tool you are using the knowledge so it knows how to act for a specific task. I think that's always going to be case. Yeah, because otherwise it's just, yeah, so like, I think this is my biggest point of view with agentic AI is like, anybody can push the buttons.

[1:11:01] It doesn't mean that the output is going to be good. You have to, this is where the, this is where the human and the AI come together. It's my brain that is inside all of these agents is my way of thinking my way and years and years and years of, you know, corporate experience years of, you know, having my own intellectual property, my critical thinking skills, analysis skills, etc, etc. Those are the things that are built into the agent that make the agent valuable, not the buttons and the things that we connected.

[1:11:35] So again. What AI model would you use for this one for example? Did you connect this to Claude or something to kind of write? Yeah, so I'm an avid Claude fan. Okay, so how much would this one cost to run every week for you as an example? So let me see. 37.2. Okay, it's pretty good. I mean, Okay.

[1:12:02] it would cost a lot more to hire someone, I guess, to do all of that. Exactly. And again, even more so when you start having it out of post, I'm currently paying manual team to do that because that's my choice. Yeah. But again, so it's like all of these agents. I'm trying to get back to the main.

[1:12:22] The other thing, and this is something that I learned from my mentor in this space, works really, really well is having all of your agents in different folders. So, agents are the ones that are like the morning briefing agent, the competitor analysis, the ones that are going out there and scouting for information and bringing it back to you.

[1:12:45] Produce agents are the ones that actually spit out an output, like the one we just looked at. Process agents are the ones that take agent output and pass it on to the next agent and then maintain agents or agents that ~ kind of just do kind of administrative ongoing maintenance type of tasks inside of the business.

[1:13:07] Okay so like those are I'm trying to think what else here we have say the LinkedIn daily connection builder... is another example. this is the one that goes and finds your ideal clients on LinkedIn. This one runs every single day. So there's no the last example is an example that the human being is the trigger because I actually have to write the newsletter.

[1:13:28] OK, this one here is just sets at 11 o'clock every day. OK. And what we have here is and this is a really good example of one that we had to troubleshoot. We had to actually break it into four different streams where we used to have like going out 20 people... all in the one agent and it kept bugging out on us and it was really frustrating so what we've actually done is staggered each of these runs so that it runs four times a day and then we were able to solve for the agent bugging out and basically it was just like there was too many steps and it just kept, it was never able to complete so we stripped it apart.

[1:14:09] So the whole idea here is that it logs onto your LinkedIn session, it goes to your LinkedIn spreadsheet, it checks the people that we've already connected with, then it goes off and scouts for more people and then connects with them, logs them on the spreadsheet. And then it also captures whenever we've sent message one, which is like, Hey, I'm growing my network and LinkedIn, I'd love to connect. And then it captures that that has been ~ sent.

[1:14:37] And then, but we've stripped it out now to run four times a day instead of trying to do 20 all in the one batch. Okay. So you'll come up against that when you're building these agents. That's, you know, the spray. And on LinkedIn, is it on terms of service or can you get your account banned if you do too much of these things?

[1:14:56] I mean, obviously they have it. Yeah, so that's why we have it set at 20. Okay, so if I try to do 100 a day, Okay, yeah. probably. ~ But again, because these agents are actually logging in through computer use, they're actually logging into my account. So whenever you initially set this thing up, you got to...

[1:15:13] you give it access to the things you log in, it your credentials, et cetera. So it's actually logging in as though it is me. And then it's adding five people, updating the spreadsheet. And then a couple of hours later, it's coming back and adding five more. So it's mimicking human behavior.

[1:15:28] Yeah. Yeah, that's probably the key on LinkedIn because they are always tricky with anything automation. But I think if you do it this way and not do too much of it, Yeah, yeah, and that's I think it could be fine. one and again with anything it's like if you choose to do it That's your choice. So this one here, Yeah.

[1:15:44] YouTube presentation coach, there's not probably too much point in popping the hood on all of these, they just literally go through the same process. They reach out to different strategy docs that we have on Google Drive. And then it's super, super simple, extracts the video, calculates all of our speaking metrics, researches best practices for YouTube videos, and then tells me, ~ here's how to do better on your YouTube videos and posted over to Slack.

[1:16:13] So it's super, super simple. That was just kind of a fun one more than anything else. It was like. ~ Yeah, I mean, this goes to show it doesn't have to be always so complex or even even. No, it doesn't need to be a 75 Yeah. step n8n like mind bending agent. it doesn't have to be, know, some of them, Yeah.

[1:16:29] and some of them need to be super complicated. I get that. But there's some, I think for people who are maybe new to this, just the only way I see so many people in paralysis, the only way to figure this out is just to start... like first of all do your design work like you're over that overview work and then just start playing with it because like all of us myself included this is new to everybody on this planet so we're we're all figuring it out you just have to pop the hood and be willing to make mistakes not get all caught up i see this a lot about people Obsessing about credits and how much it's costing it's like we literally have the world at our fingertips and we have like more opportunity than any other time

[1:17:13] in the history of the world I believe and it's like people are grumbling over a few credits. It's like we literally have so much opportunity it's like roll Yeah. up the sleeves get your hands dirty have fun with it and I don't know maybe I'm weird but that's me. No, but you've shared a lot of value here, but people or a creator would kind of get started with a few agents. I mean, you covered a lot here, but let's say, I mean, people are going to start creating content and do different things in their business.

[1:17:42] What would you recommend to people so they can maybe set up their first one right after this session? Yeah so what I would do is go to Claude, ~ please don't use ChatGPT... go to Claude and literally I use a platform called Wispr Flow ~ if you use it your life will never be the same again it is absolutely incredible.

[1:18:00] Yeah, I use it too. It's amazing. ~ It's like everybody out there like oh yeah, but I already have voice dictate and I'm like this is the best So just use Wispr Flow, which means you never have to type ever again. That entire thing that I went through today that entire presentation I started creating it at about 10 30 p.m.

[1:18:17] last night because I had a stacked week and my childcare fell through and all of the things and it was able I was able to produce that by literally talking to my computer for about an hour last night I did a little bit of refinement got up very early this morning and was able to refine it and it's just...

[1:18:33] So use Whisperflow, talk to your computer, literally say, here's my business, here's what I do, here's what is happening for me inside of marketing, here's what's happening for me inside of sales. I want you to help me to map out all of the divisions of my business and all of the interactions between each of the divisions.

[1:18:52] And what we're going to do next is identify the agents that I need to build first so that I can solve for the biggest bottlenecks that I'm struggling with. OK, so don't blow smoke and say everything great, if it's not tell it exactly where things are going wrong tell it where you're bottlenecking tell them what you need more of and then literally say to it tell me what types of agents I need to build okay so then it's going to come back and say here are these like number of agents that you could build and then say great what order should I build them in and then they're going to say do one two three four and then you say cool right let's do number one how do I do that and then it'll literally tell

[1:19:32] you how to build the agent and all you gotta do is have that curiosity, ~ that willingness to learn and the willingness to figure it out. Don't be scared of it. You can't really break anything. You might burn through a few credits as I said, but like so. ~ But go and then it's like as you then start to get into like mapping out your first agent, let's pretend it is content repurposing.

[1:19:59] It's like, OK, cool. I'm going to use the platform, Lindy, to build this agent. Create an instruction that I can give to Lindy so that it can easily and accurately build this agent without much rework inside the platform because that is gonna go through a gajillion credits. Okay, so you're then so you're using Claude as your playground to ideate to design to build your instructions and then you're only going over to Lindy to push build test and go.

[1:20:29] That's such good advice, actually. People start building right away without thinking sometimes. But I think that's really good. Yes. And ~ would you recommend, Lindy, I now we talk about the tools, the tools can change. But at this point, Thank when we are doing this session, would you recommend that if you are a beginner to AI agents, or would you try other tools in terms of cost value ratio?

[1:20:52] Yeah, super, super, super simple. ~ know, I'm probably not going to recommend jumping into n8n if you're new to this. The whole idea with these platforms and this is what we pay for to have access to them is that they have sold for and handled all of the dev work, the complexity, all of what used to take, you know, a week to build can now be done in seconds.

[1:21:16] So pay for the privilege to have access to the platform that can do all that for you. Have you tried MindPal as well? Have you experimented with that? Yes, I spent a few months on MindPal last year and then left it not for any specific reason, more just my whole piece is like we're rolling these out for clients so I'm coming at it from a different lens, not as a consumer or a user, sorry, but more so I have to have this streamlined SOP where that part of my agency are building these agents and rolling them out to clients every single day.

[1:21:54] that process as simple as possible and easy for the client who isn't tech savvy to be able to use and understand is the thing that I was optimizing for. Yeah, I think maybe the UI and stuff, maybe it's a little better in Lindy as well. I don't know if you think that. think the Mindpal, yeah.

[1:22:10] think it's just pretty basic for, know, it's like if you're a non-tech person and you're kind of scared of AI and then you get handed all these AI agents, it has to be understandable. Like nobody wants to see the brain of what's going on behind it, except for the geeks like us, you know, like we want to see that, Yeah.

[1:22:27] but the user wants it to be beautiful and the brain of it can do whatever it wants in the background, but that interface is super, super important. But, 100%. I mean, you've shared so much value and I'm sure if you are here right now and watching, you have gotten so many ideas for AI agents you can create.

[1:22:45] But I would say just go to Claude and do what Maeve shared here. I mean, go and use Wispr Flow or something to share more about your business, figure out agents you create, and then use a platform like Lindy or something. mean, it doesn't really matter what you use. Just get your feet wet with this basically.

[1:23:04] But Maeve where can people connect more with you if you want to connect Yeah, so LinkedIn on social media? is our main platform, so just ping me a note over on LinkedIn at Maeve Fergusion. Yeah, I will link that up below here and leave your biggest takeaways below as well if you have found value from this session.

[1:23:21] Thanks so much for coming on and I'll talk to you guys in the next one. for now. Bye. Okay.