I ❤️ AI Q&A: AI Automation & Metacognition w/ Examples of AI Arbitrage & How to Wake Up Cold Leads

is the, I don't know if how many of you have played around with it, but basically it's like this internal work area in AR or in VR, in Oculus land. Anyway, it's kind of cool. All right, welcome in everybody. We are now live. So thank you for waiting. Welcome to the I Love AI call. This is our chance to kind of get to know each other and hang out, but also for me to share with you some of the things that are happening in my world, in the world of AI, some of the cool conversations that I'm happening, I'm having., where I've been, where I'm going, and then most importantly, to answer some of the questions that are coming in from our own community members. So every time a new member joins the, I Love AI community, and we are, I should know this by now, but let me go look. Uh, we are, I think, close to 1,500. We're over 1,500, but I do need to like, what is happening right now? There we go. Um, we are at— oh, we're at 1,605! What? 1,600 members! Oh my gosh, I could cry right now. I did not know that. We have to do a party. Um, oh my goodness, this is so exciting. So congratulations everybody, we are 1,600. And I think this is the fastest we've hit 100. It's been less than a month, so that's a lot. But every time somebody joins the I Love, uh, AI join the community, they are asked a question like, what do you want? You know, uh, what are you looking for? Why would you, why do you, why are you even joining the community? And I wanted to use at least a little bit of our time today to talk about some of those questions and share with you some of the answers. So that being said, first I'll give you a brief recap of where I've been and where I'm going. Uh, this week and the end of last week, um, we had to postpone our call until this coming week, uh, or we basically just talked on Monday, but that was because on Friday I was speaking in New York City at the first Generate conference held, uh, it was hosted by Modev and Fuel IX, which is a, um, kind of a, a portion of TELUS Digital. Good friends in the voice world, uh, have moved and, and are doing amazing things. And so they invited me to come and do a, a keynote. In New York, and then I also got a chance to run a workshop. That workshop was pretty awesome. I was very excited. It was a brain dump of all the cool things I've learned in the last 2 years on building, uh, responsible AI at scale. Uh, it, you know, it, it was, uh, some laughed, some cried, like it was very good. I did record it, so I'll be posting that in our AI Leadership Institute. I would say like by itself, it's, I mean, People will probably pay, I mean, people did pay a lot of money to hear it. So, so I'll provide it to our, our AI Leadership Institute members. We also have a couple graduations happening with our AI Leadership Institute members, but as well as other people who have just taken some of the classes that we're offering. One of them is our Thursday Night AI Startup Accelerator, helping companies learn what it takes to grow a business and then doing the work. It's the hardest part of the accelerator is actually doing all the things. It's, A 6-week program takes longer than 6 weeks to kind of get all the stuff going, but it's great to kind of know what you need to do and then just put your hands on a keyboard and build, build, build massive action. So that, uh, class is graduating. We're actually gonna do that graduation on Monday night because I was flying from Vancouver last night, uh, and I couldn't tell if I was— I got almost rescheduled my, my flight cuz I was like, am I supposed to be on? Am I not? Anyway, so we just rescheduled it to Monday. We also have another graduation of our Maven class, Agentic AI for Beginners. Over 150 students took the class, have gone through the material. We had a dozen, um, podcast-style interviews that we did with different leaders in the space. Such a fun and engaging, uh, session. It was also 6 weeks long. So those graduates will be, will be celebrating, uh, together tomorrow. And then we will be running another cohort. Back to school in August. So when, you know, my kids go back to school, I'm gonna go back to school too and teach another cohort. But it is available on demand, um, in the AI Leadership Institute. But if you wanna like do it with people, which I think is the best way to learn, um, we have another, uh, Agentic AI for Beginners cohort launching in August. So that'll be fun. Um, so that being said, that was Friday. And then on Tuesday I had to go, or I, Got an opportunity to do the second Generate event, which was in Chicago. Got to meet up and on Friday I met with one of our members, um, Yuvika Sharma. She is a leader. She's a co-founder of the Asian Women Advancing AI, um, organization, a community that was founded while they were, both her and Padmini were members of the AI Leadership Institute. So it's super cool to watch them grow. They also just hit 500 members of their community. So super, Eve would be proud. Uh, so super exciting to kind of watch all of these communities sprout up and serve people that are looking for a place to call home. Um, but I was in Chicago. I got to meet with Padmini who's out of Iowa. Uh, but she joined me in Chicago. She spoke not only at the, um, the event that I was at, the Generate event, but she also went on same day, I think, to go and speak at the Responsible AI event being held by the AI 2030 organization. Of which I used to be a fellow, but I am no longer cuz you know, I, I try to only do those things for a year at a time. Um, but it was really, really interesting and, and would, can't wait till she has a chance to come back and tell us all about her experiences. Also this week we had one of our first, uh, member-led calls where Joe, I don't know if Joe's on the call. I can't see the participants right this second, but where Joe was, um, they, he delivered a session where he talked about how to use kind of a, an, a digital note-taking, uh, process to actually capture the information about your prompts as you're building prompts. How do you save them? How do you organize them? So that was really cool, uh, to see our community not only learn from other community members, but also realize like this is, yes, I do a bit of talking on Fridays. This is what I do. I enjoy it. I hope you all do as well. Um, but you definitely have an opportunity if there's something you have learned, something you wanna share, if you wanna test out a, a talk you're gonna give, you can definitely work with us. We'd love to give you time. We can schedule it on the calendar and 1,600 people will get a message in their inbox saying that it is something that, uh, that's available for them to come and join. If you do it with us, we'll record it for you, put it on demand. Um, so that you have a chance to really kind of build your brand that way. So if there's something, what we ask that you not do is come and try and sell something, right? Like a product you're selling or whatever. But if you're just teaching what you're learning, We would love it. Uh, love to have you, uh, join us as part of that. And I think it's awesome to teach what you learn and people learn from you, but also you learn by teaching them. It's like a beautiful flywheel virtuous cycle. Um, so I encourage you if you have ideas on that to let us know. So from Chicago, I flew back home for a few hours. Don't know if you all know, but my dad is in the hospital right now. Um, it's okay. He's okay, but it's crazy when, you know, like I'm working and then my dad gets you know, falls and has a terrible fall, has to go to the hospital. He's now in rehab, so he's just on the mend. Um, but it's just stress, you know, trying to be everything to everyone. My kids were worried about my dad. My, you know, my husband stayed home, so I traveled alone. And anyway, it was just one of those weeks. Um, but we survived because it's Friday. Friday! Um, so, uh, super glad to be here with you all. But yesterday I ended up in Vancouver delivering my last session of the Generate, um, series. There is one more Generate series happening next week in Calgary, but I will be back in Chicago next week. Um, and so yeah, yesterday it was in Vancouver. It was an amazing opportunity to meet tons of cool people. One of the members of our Agentik AI for Beginners course, Max, some of you know him, he sat on a call with us and set up his laptop with us for hours on end. Really great person. But I got to meet him in person and it was cool cuz we didn't know we would be at the same event. He didn't know until he got there that I was speaking and I didn't know he was gonna be attending. So it was awesome because I recognized him immediately cuz his little face in the little circle is the exact same face that I saw in class. So that was really cool to, uh, to see Yuvika, Padmini, Max, like all these amazing people show up at these events, participate, become leaders in these events. It's just awesome to, to be a part of. So that was my week. Um, next week I will be speaking in Chicago at DSG. This is the Distributed, um, Supply Chain Group. Uh, so they, they focus on like the distribution network, like who's distributing, um, goods across the world. And so it's a really interesting group to be a part of. I cannot wait. I'm one of the keynotes that'll speak, be speaking, um, on Wednesday, and then Wednesday night I'll be flying over to San Diego, beautiful San Diego, um, where I will have a chance to speak at the NTT conference. And I'm very excited for that. Not only do I get to do a keynote, but I also get to host a panel of experts, uh, really executives putting AI in action. And a couple of them, one is the CEO or the I think the COO and co-founder of a company called Claimable. They're relatively new. Uh, I think they started last year, late last year, but they're already doing some very interesting things. One is that they're using large language models and, and generative AI to help people. The reason it's called Claimable, it's to help people appeal claims that have been denied. And oftentimes it's very difficult. I mean, I have been on the other side of VA claims my entire adult life with my dad, and so I'm constantly battling like appeal. Appeal after appeal and I don't know what I'm doing. I definitely didn't have a large language model to help me. So I am so grateful that this service exists. I think it'll help so many people. So that's really cool. I'll get a chance to speak with them. And then I also will have the opportunity to speak to the CEO of a manufacturing organization that manufactures machinery and their journey to building autonomous machines. I think it's gonna be fascinating. So that is all at the NTT conference. You'll definitely see me promote it. If you are in San Diego, let me know. Maybe we can sneak you in as part of my entourage. Um, so definitely let me know if you're in the area. Otherwise we can just hang out at Starbucks or something, or Better Cafe. Don't judge me cuz I said Starbucks. Um, we can go to a Better Cafe as well. But that is my next week. Um, and of course I'll be home, uh, Friday cuz I always take the red-eye on Thursday, uh, as I did tonight. So I was in Vancouver last night, took the red-eye and showed up this morning at 5 in the morning, uh, and went right to work. But, uh, but I will, Yeah, definitely see you all next Friday for these sessions. I'm very excited for it. So one of the exciting things, like I said, one of the, the, well, I'll make a brief announcement. You'll start to see this in the community itself. I do want to remind you, if you're not part of the I Love AI community yet and you want to join the 1,600 people who have joined us and participate and, you know, connect and stay updated on AI and events happening all over the world, you can do that by just going to noelle.ai. Ai, that's a website, my name, Noelle Russell, .ai, and you can join us. And what you'll notice is when you do join us, you'll be asked a question. Uh, and that question, I have to look it up because I wanna make sure I get it correctly so you don't get in there and be like that. What's one topic that you would like to learn the most when it comes to AI? And we ask everyone, uh, well, I shouldn't say that. In the beginning, I didn't ask anyone any questions. So the first maybe 400 people, I didn't ask any questions. But after that I got like, ooh, this is an opportunity to like actually serve people's needs. And so I'm going to spend at least a little bit of time every week answering some of those questions. And we get a lot of questions all over across the board, but there are some themes that have begun to emerge when it comes to what people are hoping for. And I would say probably the number one word that I see, maybe there's two tied. One is automation, like what does AI automation look like? And that is one of my specialties, so I'm super excited to share that. But the other, um, I'm gonna go into just a couple more just to make sure. Yay! Oh, that's a good one. What? Um, what? I know some of these people. Um, awesome. Okay, cool. Um, oh yeah, so, and then many of them are questions around being in transition, like career transition, or being a student, and what does, like, what does that look like? Like, what are some different disciplines? What are some different roles? Um, what are different directions that people can kind of go in? So So I'm gonna, uh, take the first question first and then I'll answer a little bit about the directional, uh, the directions that I'm seeing a lot of demand in, uh, and where, I mean, nothing that I say is set in stone. Nothing that I say guarantees your success in any specific area. I just offer my own lived experience and if that's beneficial to you and hopefully just inspires you to try, uh, that is my hope, uh, in these sessions. So that first question is around like, I mean, a lot of people just type in, You know, what do you wanna know? AI automation. That's their answer. AI automation. What is it? How do I do it? And what they're typically asking as I go in and have a conversation with them in the DMs inside of our community, and I'm like, what do you mean by AI automation? Many of them are referring to the, the ability to automate human tasks with AI. Now at the AI Leadership Institute, which is the company that I run that does AI education as well as AI automation implementation, one of the things that we have focused on is what's called an AI impact assessment. And it's actually something that we're gonna give you all a chance to be part of in just about a week and a half. I'll tell you more about that at the end of the call today. But in this environment, when you're thinking about AI automations, what we think the most important step to take at the AI Leadership Institute, when we go to a client, we encourage them in a workshop, we're like, okay, first thing we're gonna do is look at how your humans are working now. This is why in many of the biggest companies in the world, they are combining HR and technology together because they're realizing that human resources, knowing the humans, understanding the humans, being empathetic to human needs and human desires, knowing what a career path is for the humans who work for you, and then building AI systems that complement every individual human that works for you at an extremely precise level. That is the future. It's entirely possible today, but it does take a leader who knows how to do that, who even knows what questions to ask. So I think automation typically starts with, well, who's doing that work now? How is that happening now? Where's the friction for that individual today? And when you do that impact assessment and look at how, how could AI impact this person's role, it gives you an early first step. Now, one of the— I had an opportunity to meet with a lot of the panelists I'll be speaking with at NTT next week. But one of them had mentioned something that I thought was really interesting, which is, you know, many times people get— and we actually talked about this in our, um, Agentic AI for Beginners Maven class that's graduating tomorrow. But this concept of like sometimes we make a lot of assumptions about how people are doing something. Sometimes even as a person who's doing the work, a lot of the processing is happening kind of inside our mind and we don't even realize that we're thinking through things in a certain way, right? We kind of like are on autopilot. How many of you have been like on autopilot at work and looked up and been like, oh, or worse on your way home? I've definitely done that where I've been like on autopilot and then I kind of like, you know, recognize what I'm doing and I'm like, oh man. And sometimes autopilot will take me 2 exits past where I need to be. Um, but it's only because like it's just Wrote like, this was before COVID but I used to do the same drive every day, twice a day, an hour in my car. It was really nuts. I can't believe I would, I spent that much time in my car actually. I don't do that today, but that concept of autopilot really needs to be almost checked a little bit, right? Like, and, and in agentic AI, we call this concept metacognition. But in humans, I think I talked about this on Monday. We, we, we really need to become aware of how we make decisions and how we are executing work and what assumptions we make and what automated things, like in our own brain, what leaps are we making in our own mind that we aren't documenting? And the benefit of that is that the more clearly we understand our human thought process, the better we'll be able to clearly articulate a machine's process, right? A machine's automated process. Oftentimes this is exposed because we will have someone document what they do. We will then apply AI to that and it will break because there was assumptions made that we didn't clearly identify, or because a step was done in a way, like we tried to replicate that step in a machine and the human didn't even understand kind of how they got to that answer because they've been doing it for 30 years, right? And they've never really had to break it down before. One thing that we are noticing with the advancements in automation, and this is like not, this is AI automation, right? Not like Automation 1.0 where we were using things like Zapier and If This Then That. We're talking now about a contextually aware automation process where now I'll give you an example. One of the, uh, projects, one of the, the offers, the services that we offer is called database reactivation.. And so we have an AI system. It's a series of LLMs that will, uh, go into a client and it will take their leads, right? A bunch of people that have reached out to them to say that they wanted to work with that company. But if you go into any company, you'll notice they have a ton of leads that they never use. A ton of people have reached out. They've got a whole database full of contacts that no one ever reaches out to. So what this AI process does is it goes through every single person and just connects with them and says, hey, It's Ali from AI Leadership Institute. I know you reached out to us a while back saying you were interested in workshops on AI. Is that still the case? And as soon as they say yes or no, if they say no, they get put in the do not contact again, or maybe wait a year unless the person said remove me from your list, and then we remove them, right? All via an automation. But what we, what if they say actually, yeah. Oh my gosh, I was just thinking about that. Oh, my board just asked me for that. They will write literally that back in the chat, and I want an LLM that's going to understand that, document it so that when I go through that contact, I can see that they said these things, document it, but then encourage them to talk a little bit more about what they need and when they need it. And as they provide that information, the LLM is actively listening, providing verification. Oh, I hear you say this and this and this. Is that right? The caller or the person in the DM says, yes, that's right, thank you. Or an SMS, yep, that's right. And once we get confirmation, say, you know what, the best thing you can do right now is talk to one of our advisors. Would you be willing to, you know, set up a time? And this AI agent, this LLM service, can now work back and forth with that person and their calendar to schedule an appropriate time to get on a call with one of our advisors. And now we're taking leads that were cold, that nobody was talking to, that you couldn't sell to, and we're turning them into not just sales calls, but warmed up, activated sales calls where they basically have said, yes, I want to be educated by you. Yes, I want a workshop. And yes, I want to talk to somebody. They're on a path to yes. And it's so much— it's like adding, you know, lighter fluid to a charcoal grill that's already on fire. Like it just ignites this energy within an organization. And that's why I love doing it. But the entire thing runs on autopilot. The entire thing, right? It, it, it runs with, I write the LLM, I monitor it. I, you know, I'm constantly kind of reviewing it, but I don't have to build that automation all the time. Like I'm using an LLM to create dynamic interactions. And so automations give us the ability now not just to automate a process, in other words, create a, a, a path to calling something automatically. It allows us to move more into this agentic workflow of how do I leverage an LLM to get contextual information about the request, to maybe use the tool use pattern, right? And go look up tools that'll satisfy that request, request those tools, connect to APIs, right? Allow it to be so much more functioning than, hey, if somebody sends an email, email them back a template that everyone else gets. Every interaction now can be customized, not only customized based on the current conversation, but customized based on anything we already know about that lead. So it's a really powerful opportunity. And remember, responsible AI practitioners, we are not doing this saying, hey, I'm Allie, a human, and I'm a robot, right? I'm disclosing like, I'm a digital assistant that's gonna work with you in order to see if there's any way we can help, right? Like many of you know, when I do keynotes, I have a presentation that I give and it says, it basically is an avatar of me called Nearly Noelle. So I immediately disclose, and when I introduce myself, or when it introduces itself, it says, I'm Nearly Noelle, a synthesized avatar here to help you, blah, blah, blah. And that is an old school design principle that we have had for many, many years. Um, so it's pretty cool stuff. Like automation is so much more fascinating. It's both and though, right? Like we're gonna have kind of just those those rote automation capabilities, but we're also now gonna be able to ingest LLMs to contextualize and automate processing in a dynamic way where I'm not specifically linearly describing that behavior. And that I think is where the future is, is that, and one, one term for this is AI arbitrage, right? Is that you can build an agent, which is what we did, As a matter of fact, let me show you my very cool website. Some of you know I'm a little bit crazy. I get ideas and I build companies like, like that. So I built this company a little bit ago. It's now going on about a month, maybe 2 months. But, but it came out of an idea where I was like, oh my gosh, what if I could just help people with this? I could build it once, customize it for everyone. So let me share that with you in— let's see. Let's see, I'm gonna go here. Hopefully I'll find it. There we go. So let me go here. These are some of our members. So I'm gonna go to Agentik. Maybe it'll populate. Oh, there we go. Agentik, aiagency.ai. So this is, uh, my very simple website. It's a templated website that I just added my copy to. I use generative AI to help me with the copy, though the idea of course came from me. Um, it's an automation agency. Let me just go to services so you can see the kind of things that we do. One, database reactivation, right? So this is exactly what I was just talking to you about, but it's LLM-driven, which is what makes it special. It does leverage Zapier, So let me— let's not get it twisted. I'm not doing too much magic, but it is LLM-driven. So every interaction is contextual. Every interaction leverages memory in the conversation. That is Automation 2.0, or maybe 3.0, 4.0. You get the idea. But I also have a Speed to Lead agent. So this, um, database reactivation agent is specifically for leads that have grown old, or what we say cold in the marketing industry, leads that you've stopped talking to. Maybe they're weeks old, months old, years old. It gives you a chance to just say, hey, it's me. Remember you said you wanted this? Do you maybe still want it? And if they do, then you got a woken up lead. But what is even more valuable to a company is when somebody sends you an email, right? Like, let's say you are, um, you've got a, uh, like me for the AI Solutions Architect Maven course. If you wanna take that course from me, And you can go in and when you take the course, there's a lead magnet that says, hey, if you wanna take this course, first thing you have to do is download the AI Solutions Architect book. And it's called like 21 Laws of AI Solutions Architecture. Well, we have about 500 people who have downloaded that book. And what I want, what I wanted is that as soon as you download the book, I, as Noelle, traveling all over Timbuktu, right? All over the world, I can't, keep up and like make sure I'm messaging everyone when they downloaded my book. So instead I created a LLM-driven listening application to listen, speed to lead, listen for when somebody downloads that app. And when that activity happens, immediately reach out to them and say, I see you, I hear you, I can answer questions for you. Would you like to schedule time on Noelle's calendar automatically? And that is hugely valuable because most CEOs, or most sales teams are extremely busy. So if we can somehow just say immediately, oh, and I don't know if you know this, but when it comes to automations, especially in sales, the reason this is so powerful is that the opportunity for you to close someone who reaches out to your company in the first 5 minutes, and actually I think it's 60 seconds, like in the first 60 seconds, your rate of closure is like in the 80, 90th percentile. Like, you are much more likely to get someone to say yes to you if you contact them within the first minute. Within the first 5 minutes, it drops significantly, but it's still pretty high. But an hour, 2 days, 2 months later, your ability to convert them drops to about 10%. And you can imagine, right? If you're looking for a solution and you go and you download someone's book, and, and you read, maybe you look at the book, maybe you don't. You go to another page. When you finally get to a page and that company calls you, you're likely gonna be like, great, I was looking for help. I'm glad you picked up the phone. Let's just do this, right? As opposed to, you know, what many of us do, which is we get these leads and we just let them sit there and we never reach out to them for many, many reasons. So that's what speed to lead is. It's just shortening the gap between when somebody says they want it, when they, and when we actually say hello. This is, um, another option for out of off— out of all hours. So if you're outside of business hours, you should have an AI agent that automatically answers the phone and more importantly says, how can I help you? Can I answer questions for you? Would you like to schedule time on Noelle's calendar? These are simple automations, but they're not, they're not kind of old school automations in the sense that they're static. These are dynamic automations Every interaction is different. It's using generative AI to create a conversation with the person during off hours. It's declaring that it's a virtual agent, a digital assistant, but it's saying, hey, but I can at least get you first thing Monday morning when people get back in the office. I can schedule you. Would you like me to do that? And so it increases your ability to convert people who contact you when, if, if it isn't, it usually goes to voicemail. And I don't know about you, but I voicemail. I might have voicemail I have not ever heard and it's still sitting in my voice inbox and they might be over 6 months old, maybe even a year. Don't judge me, it's terrible. So don't leave me a voicemail or emails for that matter. You're like, ping me on LinkedIn or SMS text me or hit me up in school. And then this one I just threw in because I'm already delivering this for a few clients, but how do we help people just get reviews on their work? Many of you know, like I always ask you at the end of a session if you thought what I did was helpful. I always ask people to provide me a testimonial. Um, and so that I do manually. Every time I talk to you all, I'm like, hey, if you liked it, can you provide a testimonial so I can tell other people that other people like me? Um, it's so much easier if you can turn it into an automation where as soon as I'm done doing something, that, that variable, that status of job complete will trigger automatically hence the word automation, automatically trigger behavior that would send a Google request. And now all of these are, uh, run in a platform known as GoHighLevel. All of our AI Leadership Institute members get access to their own, um, subaccount in GoHighLevel. So it's— and you also, if, if you happen to be doing this specific work, I have a snapshot, which is basically a cheat sheet for the automation that I just give people. So not everyone's doing this type of work, but if you are and you want that automation, which is like a lot of detailed steps, ChatGPT, Zapier integration, I have a, a package that I give to my students that just allow them to do it. Um, and, and that way you don't have to spend hours or maybe even days putting these automations together. You can just build on the shoulders of me and, and start from there. So pretty cool stuff. Um, so this is, I think, the future of automations. Does start and actually dovetails nicely into the second most common question that we've gotten this week, um, which is around like where to focus your energy if you are either in transition or if you are a student. I'll take the student one first. It's a little bit different cuz if you're a student, I have a lot of parents that are in audiences and, and they will stand up and be like, I don't know if I'm allowed to ask this, but I'm gonna ask as a parent, like, What should my kid be studying? Like, these are the questions that they ask. Um, they'll also ask me what stocks that I buy, which I try not to give that advice. Um, but definitely around schooling, I think it's really interesting and a valid concern, right, around education that if like a lot of people, a lot of kids 3 years ago like went all in on computer science and now we're seeing evidence that like, for example, Um, Amazon, right? Saved 4,500 man— people hours, uh, development hours leveraging AI to do that work. Now, what I will say is that computer science is still, I mean, I'm a computer scientist, I am an engineer at heart, but I will say that that skillset alone is probably, I mean, it, it's fine, but it's probably at more risk than a software engineering mindset combined with a domain expertise. Like, like manufacturing, supply chain, innovation, like that. How do you use computer sci— right? Computer science is basically the tool. What, what are you trying to build, right? What is the solution you wanna go after? So I was encouraged, and it also is more enjoyable in my opinion, to actually combine it with real-life practical work. So as a, if you have, if you are a computer scientist, if you're an engineer, if you're a student in engineering, it's really helpful to kind of think about, okay, If I'm in engineering, what can I, what domain skills can I complement my engineering skills with? And the benefit of that is that as the world moves forward, we're gonna see AI do more and more engineering work, right? Enter Lovable or, um, Cursor or Replit. Like there's all these platforms that are generating the outcome of code. What we have less of is people who can write test suites for that code, right? We have less people who can test, not just test it, but also be able to audit it, be able to check it for validity, be able to just do code reviews, checking for bugs. Like AI is not great at looking at its own code and calling it ugly, right? It actually thinks it writes really good code. Of course it does, but it thinks it writes really, really good code and it actually usually has about 50%, you know, 50% of the bugs. Um, are not detected by AI, uh, code review systems. So this is another area where if you are an expert in the retail space and you know software engineering, you are well positioned to be that type of leader that can help machine-generated code actually reach production in a safe, secure way. And so I do think that complement, and I think this is also true for professionals if you're in transition, is that you likely have already spent 20 years in a profession doing a thing. This is what I noticed about myself. Is that I was always drawn to solution architecture roles, which is this amazing role for some of us, terrifying for others. Um, but this amazing role that's this blend between, um, being a salesperson and being an engineer. Your job is to sit in the middle. You don't own accounts, but your job is to help the customer see the art of the possible. And in many cases, help them implement that very first POC, right? Get them going, get them excited, getting them realizing that there is like proof that this will work. Your, your job is to not only help them develop their hypothesis, but then even do their initial testing of that hypothesis. And I think that's super fun. You also then get to share those stories on stages and, and do conferences and, you know, do meetups and all of these fun things sharing the work that you're doing. So it, it's a beautiful role for someone like me who's heavily technical, really like fingers on keyboard working with the code. But also thrive in sharing what I'm learning with other humans. And there's not a lot of people that do this work well. So solutions architecture is a desperately needed role now because it is, you know, solution architecture and another role which I really like is technical program manager. Both these roles become leaders, uh, very strong leaders in the age of AI cuz solution architects are going to use AI to accelerate an understanding in the customer so that they say yes faster and buy the thing. And you will need that advocates, um, coaches, people who are going to inspire confidence and excitement around the technology. It's gonna require that level of excitement for people to actually say yes. Then the second thing, um, you know, once we have this level, like this excitement going, uh, but we also need to make sure we have someone like engineering management. Like how do we make sure that we know the problem we're gonna solve? We build that problem out in our engineering team and that someone is constantly kind of toeing the line to ensure that we're not overbuilding, right? That we're not scope creeping, that we're staying on track, under budget, that work and translating the outcomes to the business. So both of those things are very interesting directions to think about moving into, depending on your skill set. TPMs, you can be as introverted as you want. I also know lots of very successful introverted Solutions Architects because when you're in, you know, I like speaking on stage. Actually, I prefer speaking on stage in front of thousands of people. It's like my happy place. I am less comfortable one-on-one or one-on-three, 'cause I'm like, I don't know what to say. I get all nervous. But depending on, you know, what your role is as a Solutions Architect, you might be one-on-one with a CEO or an executive team. You might be one-on-a-few in front of a board, right? It's not always you in in front of 1,000 people. You don't have to be a great stage presenter. What you have to be is authentic, real, and know your stuff. And I think some of you probably fall right into that camp and have never really considered becoming a Dev Advocate or Solutions Architect, uh, before. So, so that I think is pretty interesting, uh, as well. Um, let me think. I think I covered both of them. Yeah. So if you're a student, find a domain. If you are in transition, oh, The reason I say both of those roles is that both of those roles can tie to revenue. You being the TPM, you are making the engineering work tied to a client come to life, become real. If you are a solutions architect, you're tied to revenue cuz you're getting projects across the line. You're getting customers to say yes. These are very powerful roles cuz you sit in a part of the organization that is less likely to get cut when things go badly, right? You're not gonna cut your sales team. Those are the people making money for your company. So you wanna align with the field, uh, organizations that are generating revenue or aligned with the work that's producing results, also generating revenue. Um, it's when we get stuck in these non-promotable roles, uh, doing things that are not attached to revenue, maybe cost reduction roles. I mean, that's one of the worst places you can be is in a role where we're trying to reduce costs because you yourself are a cost. And so you are already in the, in the crosshairs, as they say. Um, so amazing. Aw, thank you, Ewa. Um, so Monique was saying data will be an issue for a lot of companies simply because they're still creating documents and artifacts in silos. Yes. Uh, while trying to integrate AI, data architecture will be huge. Yes. And actually, um, for those of you who don't know, I do have, I wrote a book. It came out a couple months ago. It's called Scaling Responsible AI. And in chapter 2, I have a real, I didn't dog ear. I should probably go and note take all this so I can find it again. But in here there's a big box and it says IA before AI. And what that means is information architecture before artificial intelligence. And it kind of speaks to this idea like you have to know what your data looks like. It doesn't have to be perfect, it doesn't have to be 100% accurate. You don't have to have all your data ducks in a row, but awareness of what it is and the integrity that it has and what's missing and what you need to build and what what, what maybe you need to synthesize. That awareness is key to building a scalable solution. So absolutely data integration I think is key. As well as, you know, yeah. And I think also the world of agentic AI, which is why we're doing this Agentic AI for Beginners course on Maven, is it's so important to realize like the future is you. I mean, 10 years ago, maybe longer, I guess 10 years ago I was at Alexa. I can't believe it. So I guess 20 years ago, eek. So 20 years ago, my happy place was enterprise architecture, was literally being the engineering glue that would get two systems to talk to each other. It was called enterprise application integration, um, or E, you know, EAI. And so I, my job would be to go in and take two systems that don't naturally talk to each other and write code to get them to communicate. And I would do, there was a, a protocol back then, not unlike MCP, uh, a protocol back then that allowed me to orchestrate systems to communicate to each other, but I'd have to write snippets of Java code to get that conversation to happen. We are now in an age where that integration, that level of work that was required to get these systems to talk to each other is now gonna happen autonomously, potentially via agents, which means that the agents are gonna be smart enough if we help them, right? If we provide a list of tools and resources, if we create actions that they can use to figure out the right way to solve the problem. But agents now will have their own autonomy in figuring out what's the best way to match schemas or to connect one system to another. And as we do this more and more often, that knowledge will also become readily available to the models that we're using. So once we solve of, you know, Salesforce to SAP integration, for example, using an automation, like everyone gets the benefit of that, right? Once we know how to do that, we then can scale that knowledge across the entire enterprise. So I really love that. That is great. Awesome. So that is kind of, I always like to just spend a little bit of time talking about the questions that are coming in. We have 1,600 members, so we have about 800 questions left to go. But those are two of the ones that are the most common this week. Um, as I speak, so I'll be on two separate stages next week. Usually we get a bunch of people joining the community every time I take a stage. So feel free, every Monday we do a welcome post. We welcome everyone in. Please don't hesitate to join in and like and welcome them. Every single person in our community often started, maybe not in our community, but the everyone started not being super comfortable where they are, not knowing exactly what to do or how to do it. Um, and being in a room full of people where it's psychologically where people know where you've been and know where you could go, that hold a space for you to actually just blossom into a whole new person with brand new skills. It's really exciting. So I encourage you to join us if you wanna be on the Zoom call. We are about to break and do, um, breakout sessions. So every week we break for about 10 to 15 minutes and we just get to know each other in a small group setting in breakout rooms. If you wanna join us and realize you're not as alone as you think, you feel. Please do. Noelle Russell AI. I'm gonna stop, uh, the live session now. I hope you enjoyed it. Um, oh, before I do, we are gonna be launching on Monday. You will start to see, um, a brand new challenge. We are hosting the AI Freedom Challenge Monday, June 30th to July 4th. Of course, Independence Day for us in the United States. And if you wanna be a part of it, it's a free challenge. Every single day you'll see the ability to register. You'll also see it on LinkedIn. If you don't see it, just ping me and be like, aren't you doing a challenge? I wanna play. It's a free, basically free class. And we're gonna go from, you know, how to, how to assess where you are on the adoption curve through how to build AI solutions architectures, how to do, how to build an agent, and then how to scale. And so we're gonna have just a fun-filled week. We have speakers. It's gonna be great. So I hope you'll join us. Thank you all for being here. Happy Friday. Take some time to relax, enjoy your life, and maybe, I don't know, build something awesome this weekend. Have a great time, everyone who's on the Zoom call. We are gonna break in just a second and we will be doing a fun, um, activity where we just get to get to know each other. So I will see y'all later. Okay, I'm gonna stop the share and then I have to go in and figure out where the little button is on the Zoom. Oh, I found it.