I ❤️ AI Q&A: How to Build & Price AI Agents: Voice Bots, Lead Reactivation & the 2×2×2 Playbook

All right, here we are. Welcome in. Thank you for being here. So this is our Friday call, I Love AI. We actually are here on a Zoom where you are welcome to join in. We've already had a little fun conversation as we got started, but you can join us here. We got a good number today, but this is live broadcast on all the channels, Facebook, YouTube, Instagram Live. Let me think, what else? LinkedIn, I hope. I actually should go check and see where it's going. But typically about 8 platforms. And the hope is that what we're going to talk about today is available to everyone, accessible to everyone. What I uncovered though recently is that one thing I'm going to go back and do, which I have not done, is that when I do these, they're just called like I Love AI with Noelle. Like, it's pretty simple. But these are on YouTube, and it— and very few people are probably going to search for I Love AI with Noelle Russell, like, if they were trying to find out content from us. So we're going to go back and rename— you know, I never know what I'm going to say until I say it, but we're going to go back and rename the old sessions based on the transcript and give some idea of what's in there. So you could even pick, uh, you know, if you go back now, and I mean, we've done these now for over almost 2 years, um, if you go back, there's some really good stuff that we've talked about. You can find it if you wanted to, uh, not without listening to every single one. So we're going to rename all of these. It's also an opportunity for us to just get more visibility on YouTube and help people when they're searching for stuff actually find a really nice community to be a part of. Oh, I'm live on LinkedIn. Thank you, Renee. All right, awesome. So we're live on LinkedIn. So welcome, welcome everybody. Happy Friday. I've had a fun-filled week. I'm not going to spend too long on it because I just spoke with our AI Leadership Institute family on that. Uh, but I was in 2 cities, which is not my record. Uh, but it was interesting. Only a few times in my career have I flown to a city and then done another trip, like Inception, like inside that trip. So this time I, I headed out to MidwestCon, uh, which was phenomenal, amazing. It's my second year. I think they've been doing it for a few years. My second year there, uh, Rob Richardson and his amazing team put on a great event. It was 3 days. I actually convinced them to add a 3rd day, which was Day Zero. And on Day Zero this year, I held the first, ever AI Innovation Vibathon. And so people signed up to become part of the Vibathon. And when, as part of that Vibathon, we basically sat in a room, there's about 60 people, and we counted off by 8s and separated people into groups, and they ideated for— we only had 4 hours, so they ideated for about an hour and a half. Um, we, uh, did a few lightning sessions on how to build things, how to, uh, identify the right problem. Um, but at the end of the day, they had two options. They could either build using a tool I know and love, and Kim also knows and loves, uh, called chatbotbuilder.ai. It is the tool we use in our Agent Builder, um, 101 course. And that, uh, agent, you know, that the chatbotbuilder.ai platform was one option that they had if they happened to have gone down the road of a problem that was more conversational in nature, more assistant, coach, um, analyst type thing. Uh, but some of them wanted to build software, wanted to end up building a web app or a mobile app. And so I introduced them to Lovable, and I just like Lovable. I like the name. I mean, you can see my shirt today. I'm wearing my, uh, Lamplighters leadership shirt. So it has all 12 of the Lamplighter principles. But love is highlighted, so that's nice. I didn't notice that till just now. But every— I have 12 of these, and so every day I pick one from the closet and it's highlighted a different leadership principle. For those of you who don't know, I do have a podcast called Lamplighter, The Lamplighter Effect, and it is my leadership ethos. It's how I run my life, how I run my family, how I run my business. It's kind of fun. I'm actually leaning more into it because now I realize in the age of AI, Being a good leader is not only great when you have humans reporting to you, but it's exceptionally important when you are leading machines. And so now I realize, like, before, you know, I, of course, I'm very bullish, we'll say, on leadership skills. But in the age of AI, I'm even more so. So maybe as we have more conversations, you'll learn more about my Lamplighter leadership. You can Google it or You can find it anywhere podcasts are played. Lamplighter Leadership. I think I'm on M or P. So the first 4 leadership principles actually talk about all of these in my book, Scaling Responsible AI. But the first 4 are LAMP. Love, ambition. Well, you can see it right here. Love, ambition, mindfulness, and purpose. And I, you know, I just, I live and breathe by it. And I don't think you have to necessarily live and breathe by my leadership principles, but you should definitely have some for yourself. So spend some time this weekend thinking about like, what are those principles? What— and, and the way I came up with this was actually Jeff Bezos, right? He was the first company that was so intensely focused on leadership principles above all else, above function, where, um, they have 14 leadership principles at the time. And when they interviewed me for the role, uh, I was going to be— I was working in AI, AWS at the time. When they interviewed me for AWS, it was interesting because they were more interested in how I aligned with those leadership principles. And I honestly think they just figured they could teach me everything else. I was coming from VMware, Pivotal, Cloud Foundry. I don't know if you remember this technology, but anyway, I was coming from VMware and the Pivotal world. And yeah, so it was all behavioral around like, are you the kind of person that has a bias for action? Are you kind of the person that looks at what they're doing and tries to invent and simplify what they do every day? Are you the type of the person that is frugal? Are you the type of— you know, all these, they're asking me questions that would exhibit behaviors that align with these principles. And that set me up to be like, huh, well, I get why Jeff Bezos wants these principles, and I kind of like some of them, but what would be mine? And then I want to share with you really quickly the other thing that came to me in— while I was like developing my own leadership styles. I ended up reading a book, actually think it might be in my bookcase. It was like, how— oh yeah, here it is! Oh my gosh, how fun. Um, I don't know if I had it accessible last time when I talked about this. It's fine, he's fine. Um, uh, but, uh, it's this book. Anyone read it before? It's called How I Raised Myself from Failure to Success by Frank Betjer— Beggar— anyway, uh, Frank. So good old Frank But at the end of this book, it's, you know, it's like a little rabbit hole. At the end of this book, which is where I have this bookmark, it's like the last page of the book. At the end of this book, he talks about Benjamin Franklin's autobiography. And pardon me a second. This is like my frame of my husband. You know, I can't let it just like sit on the floor upside down. There we go. There we go. And but at the end, he talks— it's Ben Franklin and brings in Ben Franklin. This led me to read The Biography of Benjamin Franklin. And at the end of that book, it talks about how Benjamin Franklin encourages— it was really meant for his family— how he encouraged his family to improve themselves and to do that through 12 principles. Now, those are not these principles. I came up with these on my own. But he said, like, if you build your— you know, if you create a list of 12, like, principles or characteristics that you want to embody, you then can spend 12 weeks One principle every week, and you will have in a year 4 times to narrowly focus on those principles, 4 times a year, every single quarter. And I was like, that's kind of brilliant that you figure out what, what do you, what's the best version of you? What are the characteristics of that? And then actually spend time educating yourself on how to be better in that space. So what, you know, 4 times a year, 1 week a quarter, I focus on being more loving or having a more love-focused leadership style or being more ambitious and taking more action or, You know, like every single one of these. So note to self, if you, if you have principles, maybe schedule them so that you actually focus on deeply on one of those principles for a certain amount of time. For me, it's 7 days, uh, once a quarter. It serves me well, so hopefully it serves you well as well. Now, what we always do on this call is we first, um, give you a quick update on the things that I've been doing. Uh, so I was at MidwestCon. It was amazing. I ended up— my little inception trip was to Chicago. I got a chance to speak at Cottingham and Butler, a law firm in Chicago. They hired me to come and speak at their CEO summit. So I got a chance to speak to a bunch of CEOs that have this law firm as their, their law firm. It was really fun, but it was very broad. Like, there were a lot of industries, but one theme I saw in some of the industries that were very interested in the work that I did was that they were in industries like trucking, manufacturing, energy. I feel like there was another one in there that I can't remember right now, but like these— oh, home services, HVAC, roofing. And what I realized is that a theme in these industries is that the world of AI is not going to impact them the same way as the world of AI might impact like a consultant or a consultancy or even, you know, a software company, right? Or a SaaS provider. And for those of you, you know, kind of like trying to figure out like, where should I focus? Like, I really felt like the questions they were coming to me with were just so viable because they were like, they weren't like, oh my gosh, worried about job replacement. I mean, all of these organizations are seeing like pretty big impact in terms of demand for their services. You know, like that's why I love the home services side and manufacturing and housing. Oh, construction was the other one. Construction. General contracting. Like, people are not gonna stop living in houses. People are not gonna stop needing roofs and stop needing, like, air conditioning support. And so you, you know, being a provider of AI services to those types of organizations is a really interesting place to be. So anyway, I just thought it was fun. I had a great time. I got a chance to, of course, share my story, but more importantly, really connect with some of these companies. And, and I have meetings with them next week, so I'll keep you updated on what happens there. Uh, but I then returned back to MidwestCon after my vibeathon, after my time in Chicago. I'm back at MidwestCon, uh, for the closing keynote panel. I got to sit next to— and I encourage you to look them up. I didn't write all their names down, but they were very nice and I felt like we were friends. Uh, one was, uh, head of customer success at Zoom. Another one was, um, head of solution— well, I think sales engineering at Salesforce. And then the person, another person was a CEO of a multibillion-dollar healthcare company, and all of them were people of color, and it was quite inspiring. I was like, alright, let's go. So it's pretty cool, pretty cool to not be an anomaly on a stage where I typically am. There were 2 women, 2 men, like it was cool. It was cool. So, so I really appreciate, of course, Rob Richardson putting that event together, taking care and ensuring that we had a very diverse and just broad symphony of talent that was speaking on every stage. You know, it takes a lot of intention to build events like that. It's hard. It's hard work. I don't like— I don't want to do that work, but I'm so grateful for organizations like Disrupt Now that do that. And that was in Ohio. And so I got to speak with a lot of Ohio representatives, both state representatives, the senator, the mayor. It was really cool. It's always, always really fun to be part of these types of events that bring people together. Just got back today. I, I landed in, in the midst of a thunderstorm, so I got stuck on the tarmac, but just got back today and jumped right into our AI leadership call. So thank you all for who are on that call for joining me. Now what I always like to do, so that, that's like my week in review. I spoke to lots of companies and many people always like, will come up to me and ask me kind of where do I get the information that I have? How do I stay updated? I was actually speaking to someone from a financial institution, and that was their biggest question, was like, how do you not feel overwhelmed? And I just want to, like, say it out loud in case you all think that I'm, like, somehow immune to the overwhelmed feeling that you have in the world of AI. I will say I live in a constant state of overwhelm, but I also have like 6 kids, and my dad lives with me, and I have a special needs, you know, child. I also have my dad who suffers from a traumatic brain injury and is epileptic. Like, my life is pretty overwhelming without AI, so I, I maybe I attract like overwhelm. Um, so that being said, my maybe, um, aptitude or threshold for overwhelm is a bit higher, but I want to share with you some strategies that I've used to reduce the feeling of overwhelm and feel more empowered in the age of AI and Again, I think there's lots of people in the I Love AI community that also have strategies around this. So I'm gonna encourage all of y'all to share your ideas, share your strategies. If there's a thing I say that you resonate with, great, like, please let me know. But also if you have used other techniques, but I'm gonna share with you what works for me in hopes that it will help you. There's two sides of it. One is technical and the other is more personal. On the technical side, I use a couple platforms. One, I use good old-fashioned Google Alerts to send me emails about topics I care about. It reduces the signal-to-noise kind of ratio to just the narrow— well, it's not that narrow, but it reduces the amount of things that I actually get surfaced into my inbox and reminded that I care about every single day. Now, that is not AI-driven. It doesn't get better over time. It's keyword, old-school Google search type. Capability. And I also use a tool or platform called Particle, and Particle is the opposite of my Google Alerts. So I, I use them both, um, and my, my, uh, Particle, uh, account allows me to train it on what kind of news I want to see, on what kind of information I want to see. Now there's a danger to this because as I train it, it does shrink my scope of what's happening in the world to just the things I tell it I care about. And there's a danger to this. I used to work— for those of you who know and remember, I worked at NPR, National Public Radio, right during COVID It was an interesting time. COVID, a presidential election, like it was crazy, crazy times back then. But working there, we would always talk about diversity of sources and diversity of information. As a matter of fact, one of my favorite teams had built an algorithm to help make sure that when you were searching for content and sources, that that algorithm would say, hey, by the way, you might not be thinking about this in the most broad way. And so— or, you know, like, you might be going down this narrow path and only getting information that— like, reinforcement bias, right? Or confirmation bias, where it's just reinforcing this storyline you're going after, as opposed to having anything that is contrary— contrarian to that. And so one of the things that I often will do is have, like, Particle gives me a set of results, and I will ask an AI, an LLM, like ChatGPT-5, hey, here's the storylines that I'm looking after. Could you, would you please, give me some contrarian ideas? Like, I wanna make sure that I'm looking at this holistically. And I ask for that advice. I turn on web search and let it complement that., and there might be features inside Particle that allow you to do that as of this week because they launch new stuff all the time. But that's, that's how I make sure, at least theoretically, that you understand, like, hey, you can use these tools to shrink what you look at, but also use these tools to expand your purview into a space that maybe brings up perspectives you haven't either thought of or maybe are even avoiding. Yes, Jason, great race. Yes. I hope it's intentional, and I'm not—

yes, yes. So, uh, on that, so, uh, Admiral McRaven, who used to be the SOCOM commander, but when he gave your initial counseling, he gave you one freebie. So you could lose it in a meeting one time if you very passionately disagreed, and there would be no repercussions, which I think is good that you have to have that on in the real world as well, that opportunity to have a person disagree. And sometimes by that person in my case, I gave the same guidance, allowing them to disagree and like get that thing that's been on their chest for— they've been working there 5 years and just let it out. It makes life a lot easier in the future. Yeah, that initial, you know, burst is not great, but it can get a lot better.

Yes. Oh my gosh, such a good point. And so this goes back to a leadership principle at Amazon. It's called disagree and commit. And so, which is not on my list, on my leadership principles, but I highly agree with this philosophy.. And I used it all the time at Amazon and even at Microsoft, which is this idea that you can disagree and how important it is as a leader to create space for that disagreement, like to let people kind of have it out, right? Like, and say their piece. And then once they start saying, like, they give— it's kind of like you all are brainstorming about an idea and you're like, we need to find a path forward. Everyone has their chance to be like, I want this path. Oh, other— another person might come in and say, oh no, that's not a good idea. So everyone has their chance to kind of say what they think. At the end of the day, though, a leader needs to make a choice, right? Like, you say your piece. And what I love about the disagree and commit philosophy is that once that decision is made as a leader, right, once the commander makes that choice and says, we're going in this direction, you as the participant in that group, as the employee, as the vendor, you have to commit. Like, that's what the commit part is, right? You can disagree. You can say your piece. But when a decision is made, no matter what it is, you have to commit to it as if it was your own idea. And that takes a whole nother level of emotional intelligence, of like discipline. Because most people in cultures that are not really enjoyable, they will pocket veto, right? They will stick that I told you so in their pocket and just wait for the moment to be like, see, I knew it wouldn't work. See, see, I knew that was going to fail. And when you think like that, it's very hard to have long-term strategic success. And so I love that idea that, yeah, you have to give space. And the other thing at Amazon, uh, when I would have these team meetings— I did this at NPR too— we would be in like these glass rooms. It was at WeWork. I don't know if you've ever been to a WeWork where you walk by and everyone's like in a— I don't know. Anyway, I don't know that I like that, but it's like a fishbowl, they would call it, where you walk by and you could see people meeting and talking. Um, and I remember people would come up to me after meetings and they'd be like, are you guys okay? And I'm like, What do you mean? Yeah, we're okay. What do you mean? They're like, well, it looked like it got pretty heated in there. It looked like you guys were arguing. And I'm like, oh yeah, that's like diversity of thought in action. I would joke with them that like, that's what it looks like. And I don't know if you have family members or siblings, but like, I encourage my team. I'm like, let it out. Tell me how passionate you are about this. Give me the data behind why you think it's a good idea. Like, let's go. But at the end, like, even if you guys, you know, you kind of went toe to toe with somebody about your opinion, you walk out and you're like brother and sister, right? You're like, OK, all right, bro, good. Nice talking to you. See you later. Like, you're fine, right? You walk out of that meeting and you hold no resentment. You hold no grudges. You're not mad. It's not personal. But it's extremely freeing to be in an environment that's psychologically safe enough for you to just have that disagreement because you don't have to agree all the time. As a matter of fact, AI systems are better for that disagreement. And specifically in AI, we want people who are going to stretch the ideas that we have about what a solution might do. So, yeah, so I love that whole, that whole line of thinking. It's extremely important as a leader to give space to people to like think through an idea, argue about an idea, but also be that coach in the room that can like allow you to do that, but then also have that leadership discipline to say, okay, now we're gonna commit on a path forward. And whether it was your idea or not, we're gonna commit to it as if it was our own idea. And as either a leader leading that or as a participant going along with that, I have found that it's extremely valuable. And it is very hard, though, when things go the way I thought they would and I want to say, I told you so. It's very hard not to say, I told you so. So often I will say things like, I'm not going to say, I told you so. Some of you might have done that before. It's a little passive-aggressive. I wouldn't do that. I'm like, I'm not going to say it. I'm not gonna say it. I wouldn't say that, even though that's exactly what I'm saying. So anyway, one of the things that is also helpful if you are a leader in this space and you are leading teams making decisions about AI models and things are getting heated is to equip yourself with communication skills that really help you navigate these types of relationships. We talked about this before. I can't remember which call, but there is a book that you can read or listen to on Audible. It is called— Nonviolent Communication, and I don't remember who wrote it. Someone could Google it, maybe look it up, um, or grab it on Amazon. But Nonviolent Communication— there's also a Nonviolent Communication workbook. You can create a small little group of leaders and like go through it together, but it really teaches you how to navigate communication in these environments so you're not triggering people. People don't get super angry, people don't drop their badges and walk out of the room. Like It can get pretty heated, especially right now, especially in a time where we're just kind of at an elevated level of emotion just in general, certainly in the United States. Knowing and having the skills of this nonviolent communication, it's not really about violence. It's more about how do you be like endearingly empathetic to everyone in the room and their— not only their right to be heard, but also how do you navigate, right? Identify— you have to pick a path forward as a leader. How do you make sure everyone comes along with you? So Great book. I've read it multiple times in my career. Satya Nadella has recently recommended it to his C-suite. So I'd encourage you also to take a look at that. So that being said, that is one way that I kind of think about and navigate the technical side of how do I know what to pay attention to and what not to Google Alerts, Particle, and making sure that I've got either a human or AI process that can make sure I'm looking at things holistically and not in a single silo. On the personal side, uh, one of the reasons that I'm just— I can deal with, um, overwhelm a bit easier is because I adopted something called— my husband and I both, we adopted a philosophy, I guess, or a practice of, uh, voluntary— it's called voluntary exposure therapy. Some of you might have heard it before, but basically you expose yourself to doing really hard things And that way, when you're in an environment of overwhelm, that feeling of overwhelm, it either dissipates or you're just used to dealing with it. And so your nervous system is more calm, you're able to make better choices, you're able to not get emotionally engaged. Um, and so things that I do to do this, I won't spend a lot of time here, but one, I'm— we are ice bathers. We have two ice baths in our house, one upstairs, one downstairs. Uh, some, some days I'm like, I don't want to. Um, and that says a lot about who I am that day, but But it is a, it's a mental discipline, and that mental discipline, the more you exercise it, the more it serves you when you're trying to figure out where to spend your time and what to focus on. Um, and that feeling of overwhelm tends to dissipate in that, in that case. I also do just do hard things. I run half marathons dressed as a princess. That's hard. Um, I also do Bikram yoga, which is, uh, hot yoga, 26 postures, 2 breathing exercises in 40% humidity and 104-degree heat is intense. And I remember when I took my husband for the first time, he's like, I've never done anything so hard. And he's a, you know, professional soccer player, like he knows, he's, he's done hard things. And he's like, this is the hardest thing I've ever done. And it challenges you, but only 90 seconds at a time, right? Like, one of the, the founders of this yoga, he would say what you want to do is, you know, it's better to give yourself what you feel like is a heart attack in a yoga posture so you don't have a real heart attack in life. And it's kind of the same idea, like, do these hard things. Uh, oh, and I also mentioned that guy Chris Noth, I think is his name. He's a high-stakes negotiator, and he would always— he would say in his training, you want to use low-stakes environments to, to practice high-stakes conversations, high-stakes decision-making. And that's— that means For example, some of you with no brand and have never gone on the internet, like you might go on YouTube and create a video because nobody's watching. So it's a low stakes environment, but maybe you're practicing building out a solution or being a solution architect or demonstrating a product to a customer. So what can you do to like stretch yourself in a low stakes environment so that when those high stakes environments start occurring, you've done that work. You're not, you're not experiencing that for the first time. And so all of those things help me manage overwhelm, help me take action when I have so many things happening, so many new products that are coming to market, new features coming to market. Like, it can get really, really overwhelming. So, so those are the, those are the techniques that I use to manage it, uh, and hopefully, hopefully that's helpful to you. Um, but, uh, but one of the reasons why I mentioned that is because some of our— I always like to go into our community and take a look when Anyone joins I Love AI, and if you're not a member of the community, you can definitely join. You could just go to myname, noelle.ai. It's a website. It'll give you an opportunity to either join our free community, which is where this call is hosted, or you'll see an option to join the AI Leadership Institute if you're more into getting certified and fingers on keyboard and building things. But in either case, you then get to join our Zoom calls. But the first I ask you, I think, 3 questions when you join, and one of them is like, what's your burning AI question? And so the last 3 people that have joined, I'm gonna go in and just make sure, confirm, by checking out these members and make— oh, we got a, we got a couple, like, since I looked at this last, just a few minutes ago. So let me go in and see what the questions are. Right, cool. So the last 2 were on AI agents. They wanted to understand a little bit more about AI agents. And this one, which I definitely really like, Haral. I don't know if he's here. He might not. He just joined, so I don't even know if he knows we have a call on Friday. But it says he was interested in AI governance. So I'm gonna spend a little time in both. One of our AI Leadership Institute members kind of dovetailed on that AI agent conversation and was really interested in kind of the backend of AI agents. What happens if you are building a business, building those agents? How do you know about how to price them. And so I'm going to give you a couple of ideas around both. One, a quick introduction to Agentik AI, then a talk about— if right now there's an Agentik AI challenge happening in the CRM that I use in our AI Accelerator program, um, and so a bunch of, uh, people are going to start building some agents. I wanted to give you an introduction, but then also talk about like, okay, let's say you've built a few agents and those agents are specific to, let's say, the insurance agents like the insurance world, how do you go to market? How do you get that into the hands of somebody in an insurance agency so that they could use it? How do you price it? So I'm going to talk a little bit about that as well. Um, so we'll start first with some agent— yes, thanks Kristen, it was Kristen's idea, so thank you for being here, um, to hear it. Uh, and you don't have to play it, you don't have to watch it on the replay. Um, so the, the first thing, of course, is just to recap on what Agentic AI is in case— I won't say, like, in case you just haven't heard. Who knows? I talk to a lot of people. So no judgment, anyone who's like, what? Who hasn't heard of Agentic AI? It's everywhere. I will say, we tend to be in this little bubble in the AI world where we think everyone's been talking about this for a long time when really just we've been talking about it for a long time. I spoke to a bunch of CEOs, that were customers of a law firm, and these are in organizations that are not even adjacent to the tech world. So most of this was pretty new to them. And so I encourage us to have empathy for organizations that might not, and individuals who might not have actually heard. So I'm just gonna briefly overview what that means and some ways to actually practice using them in some tools that we, we use, know, and love in the AI Leadership Institute. So agentic AI is different than generative AI in the sense that we are now seeing LLMs that are not just next word predictors, right? Which is what LLMs used to be. I mean, that's kind of where they started. ChatGPT in 2022, it launched as like, hey, look at this pretty cool tool. You ask it a word, it's gonna analyze the structure of your sentence and predict what it thinks you want, and it's pretty good at it. So that LLM capability As with all things, eventually we start evolving and thinking like, wow, okay, I like that it can do that. Thank you for generating, you know, the notes from a meeting or creating action items from a transcript. That's pretty cool. But what if— here's the dream, right? Like, let's say I was talking to the Zoom people, right? So Zoom, we all got— we all, you know, got to know Zoom pretty well during COVID during those years. But let's just say we're on Zoom and it— many— I see many note takers on our call today. Those note takers are documenting the meeting and they're going to then potentially create a list of action items, right? If we're in a business meeting, maybe it says, oh, Noelle, you said you were going to do this, this, and this. You know, Kristen, you said you were going to do this, this, and this. Imagine, if you will, that not only does it take those notes, not only does it extract out action items and assign them to people, But imagine if it also had the ability to communicate to agents inside a system that would do those action items on behalf of the user. Now, I'm gonna pause there and fast forward to a moment not too far from today where organizations will be creating human complementary AI agents, meaning that if I'm a director in a consulting company, I will have an agent, and this is true inside Zoom, but I will have an agent that complements my skill set. It listens to my emails and knows what I do, knows what I think, knows my OKRs, understands my business. It's not me, but it's a reflection of me. I have trained it on me so that it can interpret certain things. And I'm going to tell it, hey, here's the threshold that I want you— if you run into situations in this threshold, send them to me. Otherwise, feel free to handle it on your own. Now, that agent shows up in a, in a meeting. Our agents end up having this meeting without us. They have a conversation. They know each other's context. They know what's happening. An agent gives me the ability to not just be a predictive LLM, but actually reach out to other tools to do the job that's being asked of at the time. So in the simplest form, it would be like the note-taker, right? Being able to call out to, you know, Noelle, CEO Noelle agent, and ask, are you available on this date? Noelle the human would like to avoid having to spend time answering that question. So Noelle the agent, my, my nearly Noelle, right, agent has the ability now to be called on by the Zoom note-taker to say, hey, Noelle agent, can, are you available this date? Can we put this time on your calendar? Can we solve this problem? And so this is— we're not far from this because there are organizations that operate like this today, is that you have the ability to create this agentic network of these systems that are not just bound to what you can do to answer a single question, but actually have a collection of tools at their disposal to do things on behalf of the user. Yeah, go ahead, Jason.

Yeah, so with the tokenization on the agents, so you can build a team and basically put a supervisor on them. So like, whatever, assign 6 people under it. And basically the leader says, nope, that's not good enough. So you'll get— it'll further refine the model. Or you can set up to be more like a symphony. So one person in charge of lots of different all at once. Or you can set it up to do jazz as well and let it be a little more random answer. But there's model and there's also like an animal model you can use as well within the agentic models. So there's lots of different ways you can make those agents either work together or not work together. There's also like an art of self-deception where you can make them all do their own independent thing and get that many answers as well.

Yeah, exactly. And these are called, uh, agentic design patterns. So these have been documented. They are not relatively new. These are also mirroring some design patterns that we used in the age of services and service-oriented architecture. Um, so tool use design pattern is one of them where you just equip an agent to have tools at its disposal. So imagine a pizzeria having the ability to create an agent that's the customer service agent that does the chatbot capability on the website. But then that agent has the ability to talk to OpenTable and book, like, tables and reservations at the restaurant, but also to Slice, which is a delivery service for delivering pizza. So that single agent now has the ability to do more, and no additional humans are required. I also think, to your point, Jason, that there is an opportunity for you to really think about this. That's why I'm focused on leadership at at the beginning of the session because you're now gonna be making these, what you would do with humans, right? You would decide how are these humans gonna work together as a leader of that team. Now you're deciding at an agentic level, how are these machines going to communicate? So understanding who's providing the input, who's providing the output, what does success look like? All of that becomes even more important in the age of agentic. AI. And we're seeing, like I said, agentic organizations where they're intentionally partnering with human resources to build out these agents to complement the humans doing the work. So like I said, I don't think we're too far from agents actually meeting and understanding what the next steps are or doing a scrum. I, I use ClickUp, which is a project management tool, and it has my favorite feature is a virtual standup. So I don't actually have to meet with anyone. I love you all though, but I don't have to meet with anyone like face to face. I could just go into my AI and into ClickUp. I can go into the project and I can ask it for the standup details and it goes and just looks at the artifact differences, the delta, the changes that have happened in all of the repos and all of the projects and all the documentation. And it summarizes that in the same way that the human would be summarizing its output. So Lots of really cool stuff. But the, the, the, the moral of the story is when you build an agent, right, you are, you're, you're building, I mean, typically they tend to be LLM driven to support the flexibility in the user interface, but they also tend to be equipped with lots of different options for getting their job done, meaning other tools, other resources, even other agents. So that being said, there's a huge market for hyper-personalized agents today. As a matter of fact, Microsoft launched in the news, Microsoft launched its capability. It's got a couple of different agent APIs that have now evolved and are available in— it's called general availability when everyone can use it. And these agentic solutions, principles, allow you to like build a custom Copilot. We've talked about this. Some of you have built it in a live learning lab with me, right? You build a custom Copilot, but that custom Copilot is now an agent because it's not just a Copilot LLM. It's an agent that has access to looking at SharePoint and talking to Teams and reporting back in a dashboard on Power BI. All these tools are inside of, like, the capabilities accessible to an agent that you built. The hardest problem will be defining what you're going to build with this technology. So Copilot Studio is one of them where you can build out an agent. And what I'm seeing in the marketplace is that customers are looking for very specific outcomes. So if they're an insurance company, they want to start with an agent that already understands the insurance agency. If they're a healthcare company, they want to know that they're working with a model that understands the difference between PH, patient health data, and PC, patient care data, and patient PII data, and HIPAA compliance. They want to make sure that the models they're using are aware of the restrictions, compliance environment, that those systems will be running in. And so there's an opportunity for us, right? So as you have lived experience in a certain area of the world, whether it's healthcare, finance, manufacturing, home services, you now can take your understanding of that and help build agents that are custom to that environment, to that industry, to that vertical. And I think— I mean, I know, like, that's, that's the future. That's the direction. We also call these small language models, right, as opposed to building big LLMs like Claude and Anthropic, and there's still places for those. We're building smaller language models where we can control the dataset that's being used. Some of you might know the term, right? RAG, Retrieval Augmented Generation. Like, we get to control how that LLM gets its answers, where it is not— it's just its training data, but actually what its inferencing data is and tell a model that its inferencing dataset is different than the data that it's been trained on and control how creative it gets with that data. So that, I think, is, so, so right now there is on GoHighLevel, it's a CRM built for small to medium-sized businesses. They're running an agentic AI challenge. Not that I'm trying to create more competition for myself. I'm not. But if you wanna do and you have a GoHighLevel account, you can go in and participate. There's only 9 days left of the challenge. There's already another, uh, you know, there are people have been doing this for— I think it's been around— the challenge has been going on for almost a month, but there's 9 days left, plenty of time to do some damage. That's my philosophy. But they're saying like, hey, help us build a marketplace full of agents for our small to medium-sized business customers. And the first thing I did is just go into the marketplace and look up where these agents are and like what kind of templates are already there, who's already published some that people can use.. And what I noticed is that in these niche categories like manufacturing, supply chain, insurance, legal, right? Like in these different areas where there's entire, you know, don't like— we call them a niche, right? Or a niche. Like there's a whole niche, a whole genre of customers in this area. There's no agents to support their desire to do something. And so now you have an opportunity if you want to go participate. You can create an account on GoHighLevel. I'll definitely send you my link so I get credit for saying that. But we create an account on GoHighLevel and you could just build them. I don't necessarily need a challenge like that, but I am quite enticed by competition. So I'll be doing it. I've already launched two of them in— I'm not even going to tell you what industries, but in two specific industries, I built a super small language model that was custom trained on that industry data. So that when the company goes, is there a— I'll just say insurance model available. Insurance, pretty broad. Yeah, Noelle Russell Russell, I agree with you. Insurance is pretty broad. So you probably would want to be like, is this a life insurance company? Is it car insurance? Is it property? Is it, you know, what is it? And then based on that, create a custom smaller model that supports it. I've talked in the AI Leadership Institute about model selection. So We won't dive into that now. But what I really wanted to get to is, okay, let's say you've built some agents. And we have an agent builder course, so you could take that. If you have questions on this, you can ask them in the community. Um, there's lots of YouTube videos on how to build an agent. So if you want to dive deeper in that, and of course, we have the Agentic AI Practitioner Certification Program in the AI Leadership Institute, which you're welcome to investigate as well. But what we do, um, oh my goodness, I hope my— yeah, it froze. No, my, uh, internet doesn't die now that all my kids are home from school. Yeah, we kind of lost you. Um. I'm back. Can you hear me?

Yay!

Zoom crashed. I'm so sorry. I don't know if you all saw that, but y'all looked great. I'm just gonna say Nobody had a weird face or nothing. It looked fine. Um, anyway, I think I have a weird face. Okay, look, I'm like checking. Okay, it's working. All right, a little bit. We'll see. I shut everything down, but no, it was Zoom. Zoom was crashing on me. So I'm gonna re— uh, oh, that's so sad. Oh, it says I'm still live, so let's see what— I'll edit it in post. I'll edit it and post, y'all. Um, all right, cool. Thank you, thank you. All right, so I don't know where I stopped, but it's somewhere around the agent challenge, right? Did y'all hear that part? All right, good. Kim, did you hear me give you a shout out in the last call? I did. All right, but you got your certificate, so congratulations. Oh, thank you very much. Yeah, I added that, so that is exciting. Okay, great. So yeah, so we read out, I think we had 15 people get certified. Doesn't mean— so we ran the AI Solution Architect course over the last, I guess it was 2 weeks ago. We ended on the 3rd of September, and we ran it for 5 days prior to that and gave everyone about a week and a half to like do their projects. And we had a lot of people take it, but not everyone can finish. And I give short time windows because that's Noelle. I will I will not be able to. I will end up not finishing something if I don't have a tight deadline. So I did that. So a lot of people are like, ah, I can't finish it that fast. So don't worry if you haven't done it. I also think yelling out people's names and saying, hey, look at the cool people who did this, is inspiring. Like, you're gonna be like, hey, I wanna be one of those people. So while I was in the middle of that, and I'll come back to you in just a second, Kristen, I promise. I am going to use this as an opportunity to yell out my— to my fam, uh, that did get their certification. Um, so let me just grab that really quick. Uh, super exciting. Uh, so quite a few— some people are still here, uh, that got their certification, so hopefully this will be exciting for them. But Patrick Stewart, um, Ali Schneider, Marcella Leon, Sandrine, um, also, uh, LaTarsha, of course. Renee Clayton, Keyshawn Carroll, Sutter, of course, Deb Perez, uh, Kristen, congratulations, Raisa Jackson, Anita Charlotte Jackson. Oh, two Jacksons. I wonder if they're related. Anyway, I never said, uh, Sombria, who actually, uh, she got— she just got her certification and she's going to be hosting our Vibathon, which all y'all get to be a part of. So we are going to combine on the 26th of September We are going to combine both the 2 o'clock and 4 o'clock calls together, and we're going to do a 4-hour vibe-a-thon on a playbook she wrote based on the stuff she learned during our class. And I was like, what better way to like bring it all together? So I'm super excited. Just have empathy, you know, because people don't always want to speak in front of a whole bunch of people. So I totally was like, we should do this. Isn't it a great idea? And she was like, uh, yeah, it'll be great. So, so, yeah, thank you for being willing.

She is actually on a different time zone.

So she— it's much later for her, so she went to bed, but, um, she will be, uh, super excited to do that. Um, Kia Little, uh, Vanja and Cynthia, Dawn, Candace's, and Kim, of course, uh, which Kimmy started it all. Uh, so congratulations everyone, you all, um, did the work. It, it's not super hard work, but it's kind of hard mentally because you have to take what you learned, you have to build a small, what I call a minimum remarkable product, and then you have to like teach it to me. Like that, that's the hard part. In every one of our certifications, the end result is usually not a multiple choice question. If you want that, you can go to Amazon or Microsoft. They're all multiple choice. You can probably get by. Who knows what you actually learned at the result? But when you have to say what you learned out loud and you have to teach it back to someone, I find it's the best way to learn. So that is the barometer for becoming certified. So those are our brand new AI, AI Solution Architect certified individuals. So congratulations, everybody. I will be creating a post and celebrating you all, but I wanted to say it out loud on our live video. And I will edit the replay so it sounds like we did this all with no disaster of technology whatsoever. So the future will never know what just happened. Only us live people will know. Anyway, so what I did wanna do before we spend some time together, I always like us to come together and just get to know each other, but also talk about a topic. And the topic we're talking about, I think, today is on the backend. Like, let's say you are an agent builder or you just decide to build some agents. What happens then? And there's two different routes. One, you could be an employee. And here's the cool thing. There are some big companies hiring people that know how to build agents. So this is why taking the AI Solution Architect class, building an agent, and then publishing that as a project on your LinkedIn profile Like, do not sleep on this for those of you who did this work, because people are actively like, who knows how to do this? And there is a very small piece of the world that knows how to do this. So you have a very specialized skill, um, you know, just don't underestimate that work that you've already done. But that being said, once you've built it, so your two choices: you can be an employee building agents for others, or you can be a business that builds agents for companies. Either is good. I have done both. I've built AI systems for companies as an employee. I've also built systems as a business owner.

I won't—

I mean, you probably guess, do I look happier now? I feel like I look happier now, but I'm not going to tell you which side is better. Um, but I will tell you, uh, that I'm going to focus on, for those of you who are thinking about, even if it's a side hustle, right, going and helping the companies in your own neighborhood, Main Street USA, just live one day longer, get more customers by helping them with integrating AI into their business. So the question that was presented in our last call was like, okay, how does that work? Like, how do I productize these? So there's a couple of options. So again, I apologize for those of you who are intrapreneurs and building these inside companies. Today we're gonna focus on— but I think it's valuable for any and everyone. I've been— I always learn through the entrepreneurial lens, even if I was working as an employee, and then I use those philosophies inside my company., and it served me well. So I hope it will serve you as well. So let's say you've built a couple of agents, customer service agents. Maybe you built an agent to help with help desk or ticketing. I mean, honestly, any problem you can think of, any friction point in a process that a human has inside a company, you can build an agent for. And that's where you should start. Super small, super iterative. But you've built it now. You think it's pretty cool. And I— you want to talk to some people. Now, all— one thing that I shared with my AI Leadership Institute is that you'd want to build like a lead magnet. And I won't give you the details because I, I save that for those people, but you'd want to build a lead magnet like a some kind of document or ebook or a web class that says, here's why you need this. If you're an insurance company, here's why you need to do this. They raise their hand and now you're presented with an opportunity to say, great, I can sell this to you. How do you do that? So I'm going to give you one option, which is to start with what's called a voice AI system, which is basically every small business— like, let's say you're a small to medium-sized business— they either have a human doing this or no one's doing it. Who's answering their phone? Like, they still have phones, by the way. I know we all live in this little AI bubble and we don't think people have real phones, but companies have phone numbers. It's on their website. Someone's calling that number. Usually someone has to be paid to do that work. So what the first thing you can do is just say, How can I level up the value of that person that you have answering the phone from answering the phone and doing something else? Like, how do they become sellers? How do they add more value to the company? That voice agent— I use GoHighLevel, but you can use Copilot Studio, you can use AgentForce in Salesforce. Like, all, all of the agent platforms will give you the ability to do this as long as they were voice-enabled. But creating a voice AI agent first The reason I like it is that the scope is small. It's answering the phone. So you already know, the customer already knows what kind of questions you get when somebody answers the phone. But the most important thing that even humans don't do well is turning every phone call into a lead. Every time somebody calls that number, you're like, hey, so nice to meet you. In order to serve you well, can you give me your name, phone number, and email address? Before they even ask their question. Like, and now you have a lead that you can follow up with, that you can say, hey, I'd love to serve you at a higher level.. But the first most important thing, and so I, for example, you all can go to NoelleRussell.ai or AILEadershipInstitute.com. At the very top, you'll see I have an AI assistant that answers my phone in all 3 of my businesses. And I encourage you to call. You just call it just to see it. And you, you're already in my funnel, just so y'all know. If you don't know, you are. You're already in my funnel. You had to give me your email address to be here. So you're already in there, so you could just talk to it and see what it does and how easy the conversation is. And you could even, you know, I always say break your bot, right? You can even try and like get it to do something it's not supposed to do if you want. The conversation will abruptly stop at 5 minutes, and that's a setting you control. Like, how long do you want the conversation to go on for? 5 is a lot, but I'm like I said, it's a demo. The nice thing about building an AI voice bot is that I can call a client or email a client and say, hey, if you want to see what I can do for you, just call my number. And they can just call your number. It's the easiest sell. The other thing I like about GoHighLevel as a platform is that this service, you can build unlimited numbers of voice bots for your clients for like $100 a month. So this becomes a very low-hanging fruit for you to offer to a client. And what I do is I offer for $100, $100 a month. Now, granted, I get to build a bunch of them for that, but I'm going to charge $100 a month and $100 a month is going to answer the phone easily. They are spending more than $100 a month either missing opportunity or they're paying a human to do it. And that's— they're paying that human more than $100 a month for sure. So it's an instant way to kind of be like, let's do this as a first round. And its only job is to collect leads. And collect the answer. They don't even have to— the, the AI bot doesn't even have to answer questions. It should, but it doesn't have to. You can delegate all of it to someone will call you back answering service, but it's an AI system that gets smart with time. So first, so I guess the moral of that story is identify a use case that you can deliver super cheap but is valuable to the business, like in this case, 10x value, right? Like, I'm gonna Maybe I'm paying someone, what, $1,000 a month to be super conservative to answer my phone. And now I'm going to come in and be like, well, for $100 a month, what can we get that person who was answering the phone? What can we do now? Like, maybe they can focus on everybody who calls and gives lead information. They're reaching out to those people and selling them something, right? Like their value to the— it's not like you get rid of that person. You get to just lift up what that person gets to do for your company to do much more valuable things. So that's how I start, is iterative, small. We call it minimum remarkable product. A voice AI system is a minimum remarkable product. Now, Kim, I'm watching you. Are you calling my number? Is it working? Are you calling somebody else? Okay. So it's really fun. I'll get a text in a minute being like, Kim is calling you. So that is one, one way to, was one way to start. It's my favorite way to start. It's super low cost. Now, a little bit more fancy. Is that you create an AI agent that actually nurtures those leads. So let's say a human answers the phone, gets a phone number, puts it into the CRM. Then what happens? Usually most companies, maybe they call that person back, email or DM that person that day. But after 48 hours, they probably will never call that person again if they didn't buy or engage with the salesperson. Right. And so after 48 hours, all these leads go what we call cold.. And this is a huge opportunity for you because a company has invested in building these leads. They've gotten these leads into their database and they're cold, meaning they haven't sold them. So what I do is I create an agent that's going to warm up those leads. We call it database reactivation, and you can Google it. There's lots of YouTube videos on how to do this. It's not rocket science. What's— and some of you might wonder, like, how do, how do you make money on this? But I will tell you, this is my favorite use case because they've already got leads that no one is using. And so if you can create an LLM that simply says, hey, I noticed you were interested in this, are you still interested? That's it. 30% of those leads are going to be like, I mean, they signed up for the thing. They joined the list. Something in them said they were interested. So build an agent that just reaches out and says, are you still interested? And for 30% of those leads, they're going to be like, Actually, yes. And now you want an LLM that you don't want a human jumping in right then because maybe what you have doesn't match what they want. So LLM asks, hey, so great, glad to hear that. What are you looking for? And now you have the chance to use an LLM to qualify the lead and score the lead. And if it's a high, like a lead that's basically like, I want to buy right now, you text the manager or the leader or the company owner and you're like, get on the phone. This person's ready to buy. And they will, and then they'll close. All from a lead that they had given up on. So what I do in my business is we do all of that for free. We build the LLM, we set it up, we cover the SMS call costs, we cover the email costs, we cover all of that. And we just ask for a very small, slightly high percentage of revenue on the back end. We'll say, hey, you were never gonna get money from this. I'm gonna now close it. Your average cost of your product is $1,000. You have a 90% margin on that product because you're a consultant. I'm gonna ask for 50% of that revenue. And the first thing, they will be like, 50%, that's a lot. And I'll be like, anyone know where I'm going with this? Like, 50% of $1,000 is better than 50% of nothing. And I will tell you though, I say this I am a wimp. So people are like, 50%? I'm like, well, maybe. I mean, 35% or just being candid with y'all, like transparently. I'm like, well, let me help you help me. Like, so usually I don't end up at 50, but I start at 50 and I absolutely could get 50 if I was like a hard-nosed seller and was like, no, I'm easily going to make you double what you would. I mean, I'm taking nothing and turning it into something. You should pay me 50%. But I have empathy for people, so I'm like, okay, all right, fine. What should we— anyway, so I end up around 35% realistically, but still 35% on a model that now it's, it's called arbitrage. I build it once, I implement that in every company, especially if I build for insurance. I can put it in every insurance provider with just a couple tweaks. If I build it for healthcare, if I build it for a med spa or a roofer or a painting company, These are all like— that's why I love home services. Like, those are services that are going to keep going no matter what happens with AGI. Like, we get super smart robots, like, there's still going to be roofers, there's still going to be people fixing our AC, there's still going to be people landscaping. I was the keynote at the Landscaper Associations of America. Yeah, it's a thing. They get together and talk. There was like 40,000 people there. I was shocked. Landscaper— National Landscapers Association. Did they have AI?

They did.

They were using autonomous mowers, but not by itself. Like, you know, everything's managed by humans. Their biggest struggle and friction was actually an agent that would help them communicate between the field workers, which were Hispanic mostly, and the office workers, which were not Spanish speaking. And so they needed some— something in the middle to do real-time translation. The other thing was workplace safety. How could they, like, take a picture of a yard and know where the power was and where, right, where things were? Because sometimes they put little sticks in, but sometimes they either don't or they get pulled up or whatever. So, so a bunch of, like, so easy. Like, these are the types of things. Those are really cool use cases. But it starts with, can I just answer your phone for you for $100 a month? If you build an agent, I like the idea of just getting backend sales, right? So I'll do it for free, but you're going to pay me 35% on the backend. From a price, so you get some ideas on pricing. Classic AI conversational app building pricing is between, I would say between $2,500 and $5,000 to build an app because you're building something custom to them. But then you have to add on AI red teaming is what I call it, but you have to add on maintenance. So then it's $500 a month, $1,000 a month, depends on the scope and the size of the business. But you become the shepherd of that AI model, right? You take your cute little baby tiger, and your job is to care for that baby tiger as it grows up in that company. Oh my gosh, I'm gonna share this with you. So my way of going to market now, it's called 2x2x2, and I'm gonna encourage some of you who are in this world that this— you could do this. So I spend 2 hours with the client. This is usually free where I just educate them, right?

Arc.

Awareness, readiness, competitive advantage. So I spend 2 hours with them giving them awareness, educate them sometimes for free. Sometimes they pay me to do a keynote. But that part is like, how do I just get them educated so they know what I'm offering them? Then the next 2— so that's 2 hours. Then I spend 2 days with them, which is like what I did. I vibathon or build something with them. But the idea is to build a minimum remarkable product on the other side., right? So we got this minimum marketable product. Now, like, ideally, right, at the end of 2 days, I've got— it's like vibe coding. Like, it's, it's good, but it's not secure, it's not accurate. Like, I haven't done any work. Like, it's just, it's just a baby tiger. Like, I literally have built them a baby tiger, and they're like, I love it. Um, and then, so now I need to harden that into like a real project that has real legs. So that's a 2-week engagement. So now I have 2 weeks, and I've sold multiple of these. This is a pretty new framework, but I've got a 2-week engagement where I now harden that into something that's now ready to take on real users. Then I have a 2-month engagement where I take the real users and turn it into a production pilot. Now it's using production data. It's got its own little segmented security. I'm implementing the harder parts of that solution. At the end of 2 months, and I know some people are gonna argue with me on this, I have not. Again, if you're building incremental value, these models are not massive. You're not boiling the ocean and creating Bloomberg AI. Like, this is a single problem you're solving. In 2 months, you've got a pretty good model that is secure, accurate, trustworthy, guardrails. Now they're gonna need red teaming. They're gonna need a team that cares and feeds that model over time. And this is what I think is where the world— I mean, some of us are gonna do very well. Which is managed service provisioning. Like, we can come in and become the managed service provider for just maintaining these Baby Tigers over time. And managed service providers— I don't know if you know this— MSPs, there's a whole world of them. They're the ones that you hire to handle all the nonsense of like password resets and setting up your like SharePoint and like what happens when you get locked out of your system. Like, companies hire these MSPs to handle the nonsense that's involved in just running IT. The same thing is going to be true and is true about AI. There's just like manual labor associated with managing and protecting and guardrailing these models over time. And so now you have an opportunity to get some upfront money, right, in the process of building it. But once it goes live, you can actually just set up, you know, $500 a user, $200 a user, $47 a user. And these companies that have 100,000, you know, employees, Right? Like, that's how MSPs make their money. It's small amounts, but spread across all the users of the system ends up turning into quite a big amount. And the whole purpose of that— like, I, I sign it for 2 years. That's why 2x2x2, right? 2 hours converts to 2 days. We build something, converts to 2 weeks. We harden it. 2 months, we're ready for production. And now you sign a 2-year agreement where I help you maintain that over time. Like, it's kind of— it kind of works. Like, just try it. What's the worst thing happen? You could also do 3x3x3, for goodness sakes, or 4. You get the idea. But it's easy because it allows a customer to, like, future pace themselves, right? See them 2 years from now and being like, oh yeah, I'm going to have you maintaining this system, or at least I know I have that as an option. So when you present it, when you start with your awareness, you can tell them, like, the next logical step is the next 2, is our 2 days together, is our 2 weeks together, our 2 months together. Like, you're constantly seeding your next engagement. So for you consultants out here, like, is kind of brilliant. Like, it works. People want to know where they're going. They want you to prescribe what the solution is going to look like. Um, and so this has been working really well for us, so I wanted to share with you all too. Some of you are in this space, some of you aren't, but hopefully you'll find fun. And also, when someone presents you with a 2x2x2, I'd be like, that's pretty cool. I appreciate what you're doing. I like you helping me see what the future looks like. And at every stage, just me, ethically, I always provide deliverables that allow that that customer to take what I've done and do something else. Like, they can do it themselves, they can find somebody else. The burden is on me to offer operationally excellent service that they're like, there is no one else. And I've had people leave me and, uh, go and try and find other people that do what I do, and then they come back. So don't worry if people like go, all right, we're gonna, we're gonna go shop around. I'm like, good, yeah, shop around, have a good time, good luck. Good luck. I wish you the best. I'll even help you. And I would tell people when I was working at consulting company, I'd be like, there's probably— I can count on one hand how many people can do what I'm about to walk you through, but go ahead. And then they would come back like 10 days, 3 weeks later, the account rep would be like, so, so who were these other 4? And I was like, yeah, yeah, I actually just meant me. There really aren't that many people. But that was before I built this. So now I have you like in the AI Leadership Institute. Like, once you go through the, uh, startup accelerator, you build a company. Like, now those people are my people, right? So now I do have more people, um, but, uh, but it's kind of a flex. Like, encourage them. Go out and find somebody who can do what you do in the way that you do it. Now, I'm saying that hoping that you follow what I do in, like, my book, right? Like, that, that you build safely and responsibly, that you implement guardrails, that you infuse, like, responsible, secure, accurate development into everything that you do. If not, you're just like everybody else, and that's fine too. No judgment. Like, go on. I prefer you build safely and responsibly though, if you're gonna do it. All right, so we are a little over because I had a little snafu. Um, I would love for us to get to know each other. If you've got some time to hang out, please do. Uh, if you've got to go, I understand that as well. But we just talked about what it takes to kind of productize a model, productize an idea. So if you're in a company, you still have that effort that's happening. It would be great to hear what your experiences are. If you've ever productized a model, hopefully you can share those experiences with the people in the group, or if you're about to or thinking about doing it, just share what you're thinking, what you're concerned about, and let the group kind of, you know, talk, speak into you about what they've experienced or what maybe they've seen in their industry. Share some of the, the live learnings as we go. So we're gonna spend about 10 minutes in small groups. But while we're doing that, and I'm forming these little breakout rooms, if you have a question or comment or anything you want to throw into the chat, and of course I do have a documented list. So I do have an actual Google Doc with all of the what I just talked about with a little bit more detail. So that's in I Love AI. Sherilyn, I think I missed her. She had to leave. So I will write down a note to remember to post that. Or one of you can be nice and just be like, so, um, that thing you said, can you post it? I'm human. Worse, I'm like an ADHD human. So just have empathy and don't feel bad pinging me if you're like, you said this thing, tell me more, because I'm always happy to do it. Okay, so we've got quite a few people in here. We do have some note takers, so I'm going to try my hardest. If you end up in a room by yourself with the note taker, come back to the main room and I will either reassign you or we'll just hang out. So, um, all right, we've got— let's do this. I'm gonna go this way. All right, I'm almost there. Let's see, I'm gonna create this, this, this. All right, let me move this. There we go. 1, 2, 3. All right, cool. I'm gonna go up there. 1, 2, 3. 1, 2, 3. Hopefully you all are going to be in there. Otherwise it should be fine. Oh, Sherry's on the phone! What's up? And Joe and—

oh, what? Hi.

Okay, cool. Note takers. Oh, I want to get you out of there. 2 room. Oh my gosh, there's a lot of people in this room. 6. All right, let me move. I have to move some bots to room 5. Room 5 is the bot room. And Noelle has 2 note takers. Shame on her. Room 5. No taker. No taker. All right, I think we're in. I think we're in business. Oh, one more note taker. All right, room 5. There we go. All right, so I'm about to open the rooms. We're gonna spend 10 minutes. Don't forget to give each other a chance to say a 1-minute opening about who you are. I always call it your I help statement, but I've now, uh, become humble enough to realize it might be a help me statement. So just say who you are, uh, what you do, uh, and either what you have, what you need help with, or what you help other people with. And it just helps us understand like, who do we wanna maybe set up a one-on-one with? Who do we wanna connect with? So I'm going to open up the rooms now. Go have some fun. Make sure everyone gets a chance to talk. And I'm going to just let you know that the voice of Noelle is going to come over the airwaves about 8 minutes from now and let you know we have 2 minutes left. So have some fun together. It'll be fun. And I'll see you in about 10 minutes. Hello, Sky. I think I put you in a room. Hi, Eve. I thought you— Yeah. I can't go to a room because I have to— I have to jump. Oh, no worries. Sorry. No worries. Yeah. I have to jump, so I didn't wanna be in a room. Are we gonna find some time to connect?

Yeah.

For sure. I just got your message. I just landed today. So I just got your message when I was in— You're always just landing.

It's amazing.

Yeah. It is. But I'm home next week. So, uh, I'll send you a couple times. Yeah, that'd be great because I'm meeting with, um, I'm meeting with him in person on Wednesday. Oh, perfect. Yeah, so Monday would be good to give you a couple days to like— Thank you, Noelle. I really appreciate it. And congratulations. If there's something involved, you know, you'll tell me. Um, yeah, I can't, I can't believe it happened. I can. I got so freaking lucky. Anyway, thank you. Amazing. You're welcome. So yeah, so let's talk Monday. Hey, Noelle. Well, I was by myself, so I— Oh, good. Come on back. We could just hang out here, have our own little group. Okay. Hey, Noelle. Hi.

Hi.

I would like to connect with you one-on-one when you will have time. Can you ping Sherilyn? She has my calendar. And so I'm free. Sherilyn? Yeah. So I'm not traveling next week, so I have a little bit— like, this week, of course, was crazy, but next week I'm— Sure. I will let her know. Thank you. Yeah, you're welcome. All right, I'm going to try and put— I think I saw someone. Oh, Sandrine. So I'm going to put— Sandrine, are you here? No, I'm going to put you in room 6. She's like by herself. Poor thing. Room 6. There we go. Hopefully she'll go. But I know. All right, so I'm going to quickly talk to our LinkedIn and Facebook audiences. Let's see. Do, do, do, do. So, so nice to have you all here. We are now in breakout. So if you want to be part of the breakout, do not hesitate to head on over noelle.ai, join I Love AI community, and you'll get a chance to be here, get to meet other people, understand what they're doing. I hope you enjoyed today's session. If there's any topic that you want to learn more about, join the I Love AI community, ask your question, and we cover them every single Friday unless I'm traveling or working, and then we cover them one other day during the week. So next week I will be on Friday. I will be at the Women of Steel, I think, conference. And so I'm super excited about that. I'll be in Naples, Florida. So we will be doing our calls on Thursday instead of Friday next week. So hopefully we'll see you then. Thanks for joining. I hope you found today's content interesting and insightful. And most importantly, if you are building AI systems, I hope