Generative AI for Enterprise with Noelle Russell of AI Leadership Institute

Noelle Russell, founder of the AI Leadership Institute, shares her experience in AI, highlighting responsible AI, challenges in implementation, and the potential of generative AI in customer service. She emphasizes collaboration, flexibility, and aligning values with partners. Russell discusses harnessing team knowledge with LLMs, the role of certifications, AI education programs, and the importance of learning-by-doing and supportive communities.

Noelle Russell Quotes
🚀 "In the realm of AI, the leap from prototype to product is where true innovation takes flight."
🤖 "Deploying AI responsibly is not just an option; it's a necessity for future-proofing our technological advancements."
💡 "Generative AI isn't just about creating content; it's about unlocking new possibilities for human-machine collaboration."
🌐 "Aligning AI projects with core values isn't just good ethics; it's smart business."

Resources
iHeartAI Community: A platform for AI enthusiasts to join, learn, participate in challenges
AI Education Programs: Stanford Foundation Model Scholars Program, MIT CSAIL, Harvard GenAI
edX: An online platform offering accessible learning resources from various universities
LinkedIn Learning: For obtaining certifications and continuous education in AI.

Noelle Russell Bio
Noelle Russell is a multi-award-winning technologist with an entrepreneurial spirit who specializes in helping companies with data, cloud, conversational AI, Generative AI, and LLMs. She has led teams at NPR, Microsoft, IBM, AWS and Amazon Alexa, and is a consistent champion for Data and AI literacy. In the last year, she was awarded the Microsoft Most Valuable Professional (MVP) award for Artificial Intelligence (for the 3rd year) as well as VentureBeat’s Women in AI Responsibility and Ethics award. She is the host of the Good Morning, AI Podcast that can be heard on Spotify, Apple Podcasts and IHeartRadio.

Connect with Noelle:
LinkedIn
AI Leadership Institute

Chapters

• 0:00 Introduction and Background
• 2:36 Projects in Customer Service
• 4:11 Building AI Responsibly
• 5:03 Moving from Day One to Day Two
• 5:53 Productizing Generative AI
• 7:47 Overcoming AI Implementation Challenges
• 9:44 Improving Customer Service in Contact Centers
• 10:35 Shifting from Cost Cutting to Revenue Growth
• 13:24 Choosing the Right Partners and Flexibility
• 16:32 Harnessing Team Knowledge
• 20:06 Addressing the Aging Workforce Challenge with LLMs
• 27:11 Noelle's Career Journey Launching Early Voice AI at Amazon Alexa, "Star Trek Computer"
• 38:23 AI Education Programs and Communities

Transcript

0:00

Hi, my name is Joan Palmiter Bajorek. I'm on a mission to decrease fluffy hype and talk about the people actually building in AI. Anyone can build in AI, including you. Whether you're terrified or excited, there's been no better time than today to dive in. Now is the time to be curious and future-proof your career, and ultimately, your income. This podcast isn't about white dudes patting themselves on the back. This is about you and me. and all the paths into cool projects around the world. So what's next on your AI roadmap? Let's figure it out together. You ready? This is Your AI Roadmap, the podcast. Hey folks, this is Joan dropping by to say hello, giving a little bit of context about this episode. So I met Noelle in 2019 at this awesome conference and I heard her speak when she was working at several companies ago. I was so impressed with her presence, with her confidence, with her charisma. I'm just so in awe of Noelle and the career she's built. and all these cool companies she's worked at, and really thinking about getting your hands messy, playing with the code. Noelle is someone who is truly making the AI field better and doing the work. She talks in this episode about customer service bots. She talks about generative AI. I think you're going to love this episode. Are you ready? Let's dive Hey, Noelle. Good to see you too. Well, could you introduce yourself please for our listeners? Absolutely. I am Noelle Russell. I am the founder of the AI Leadership Institute. And I have spent the last, I don't know, decade or so really focused on executive audiences, trying to teach them a little bit more about what to do in this incredibly emerging space. And of course, just as you know, in the last year, all about AI. But we've really been talking about this for quite some time. I'm actually coming from a year at Accenture where I helped launch the generative AI practice. and a lot of those fun experiences about taking it from idea to fruition, to productizing that I think we'll have an opportunity to talk a little bit about today. So thanks for having me. Oh, so delighted. Well, someone who has such vast experience and leadership stuff. Well, I'd love to start with kind of projects you're working on, then talk about the future of your parts of the field, and then how you got into AI so it can get really practical for listeners who would wanna pursue this path. Can we talk though about like current projects as much as you're willing to share? Would love to hear. Absolutely. So last year was super fun for me because I am one of those entrepreneurial people. I always call myself an intrapreneur because I work inside of big companies like Amazon, Microsoft, IBM, but I'm always part of these startup endeavors, right? These small teams that end up turning into big teams. And I was part of the team that built Generative AI last year, built this practice and some of those projects, and I actually think it's a bit of a, maybe a secret to the success that we had. was that we were planting seeds that now are starting to harvest, right? And those projects, 80% of the projects I did last year were often around customer service. And the reason for that is because it's pretty low hanging fruit. Like one of the things, you and I have been involved since chat bots started to be a thing and they weren't awesome. Like they were okay, but they were not fantastic at the job. And so getting customers to have an opportunity um, deflect some of the questions that can be answered to a digital agent was something every company wanted, but very few companies successfully achieved in the chat bot age. And so last year I spent most of my time going into companies that already had this concept of we wish we had a good chat bot, we wish we had something that could do this deflection, um, and really improving that, right? Going in and saying, Hey, you know what? a generative model, a GPT-4, a 3.5 turbo, a fine-tuned model could do for you when you can add a data source that you own and control and manage the ground truth of. This created a lot of accelerated interest. And so today, the first thing we often did was just build, and actually it goes back to the number one question I would get from executives when I talked to them over the last year, they wouldn't ask me, Like, what is the thing I should do? Where should we start in the company? Their very first question was, how do I not end up in the news? Which I thought was fun and exciting, but they're like, how do I not end up on CBS or CNBC being that company that inadvertently harnessed enthusiasm and went to market with something without really understanding how to build it responsibly and in an inclusive way? And of course we know there are many stories of these companies now. Samsung was an early entrance, right, into this space. But a lot, it's like Buzzfeed. I mean, a lot of companies got super excited and just was like, we can do this right now. Because it seems, you know, AI, with ChatGPT, it made it seem very easy. And what, over the last year, a lot of companies have realized is, actually it's not that easy. And I've come to coin this term like day one and day two. Last year was all day ones, like, hey, this is exciting. tip of the hype cycle, people so interested. But I started to really see our executives saying, okay, great, I love it. How do we do this safely? How do we build a sandbox for innovation? So most of my projects ended up being sandboxes for testing out this technology as opposed to really pushing to production. Over the last three months though, now my journey has shifted much more deeply into, okay, we've got a POC that works, lots of people do. How do you actually productize that? And the difference between what you see in a POC with a generative model and how you get it to production, how many times you hit the model live, where latency is, caching, like security, mitigating bias. None of these questions happened really last year. Everyone was just so excited to test it out. And so a lot of, I've got a very specific project I'm working on, we're actually in phase three of this project now. But you could just imagine a lot of times over the last few years, companies, and this company in particular, was paying another company, a software as a service company, to write content for them. And in some cases, it was humans that were doing the writing, in some cases, it was RPA type processes that were automating content creation. And we went in there and we're like, all right, I think we could, not only could we do better, we're not gonna even say that, we could do the same, but instead of for just, one eighth of your audience, we would ask them, what's your total audience? Our total audience is 18 million people. Okay, what if we could do it for 18 million people instead of for half of what you're paying to reach an eighth of your population? And I love the reaction, because people will always be like, I mean, I'll believe it when I see it. I'm like, great, give me a month. And that's... Hehehehehehe Me and like 40 of my friends built this system. It successfully did that work creating content for 18 million people. We've now moved on. And here's the fun thing that happens with Gen. AI projects, right? Is that when you go in and you do something for one little team and you make their lives better, they become champions across the org. So the next meeting we go in, it wasn't just them. It was like them plus their friends in streaming, plus their friends in digital, plus their friends in. Um, CPG or product goods, like a bunch of people that we did not expect showed up and went like, what did you do here and how did it happen and how do we get some? And that I think is like, if you do it well, which is the critical part, which is why I love this podcast, if you do it well, you will be very successful, but you are coming on the heels of 10 years of kind of bad projects, projects that didn't actually deliver on the promise. So there's a lot of work we have to do, and there's not a whole lot of people that know how to do it. And so I always say, like, if you have even the tiniest interest in this space, it's a great opportunity to dive in, because if you care about a problem, this technology will help you solve that problem better, faster than any other technology I've used in the past. Wow, well, and I love the concept of like the baggage of the bad projects. You know, people, I definitely have experienced customers who, you know, the trust has just really significantly dropped. But the problems still exist. So yeah. of the sad thing, right? And I always tell people, or I always ask people like, in my mind, I'm thinking, who hurt you? Because they are so like, I'm not sure that, I mean, it's blatant. I can put it on a page, I show the numbers, like it's a very easy decision. And they're still like, I just don't know. And I'm like, this is because you had a project that looked good on paper and you never got there. And so now I'm a big believer and I encourage people to do this. I always say I go from the boardroom to the whiteboard to the keyboard. And I try to do that in a single conversation. And what that means is that your habits as an entrepreneur, as a business leader to tell versus show, you're going to have to like go against what you're familiar with, right? Rather than having a slide deck, rather than spending weeks trying to prepare a 20 slide deck to pitch, you're going to want to build something that works. that can show someone that it's different. And I find that those people that have the skills to understand the business, architect a quick solution and build something you can demonstrate, now can get past that kind of, like you said, the distrust that has occurred over the last few years. And let's just face it, it's way more fun to show than to tell. So I love it. I love that we're in a world now that requires us to be a little bit more practical. in the promises that we make. And I think it makes everyone better. It's hard, I get it, it's harder. It requires us to be more diligent in our learning and our curiosity, but it's much better experience for everyone. Yeah, I love that. And so, as you mentioned, so practical. I'm hearing this like boardroom whiteboard to the keyboard. I mean, that is like very specific and how to operationalize. Like it's a recipe, like I literally hear, yeah. literally the roadmap. I mean, no one's given it to you. Pick it up, pick it up. Yeah, that's fantastic. Well, okay, 80% of folks want customer service stuff. Can you give us a specific example or even like an abstract concrete steps? Would love to hear. Yes. And oftentimes when I'm talking to these organizations, very specifically a company in healthcare, they're looking at their existing contact center. And even early last year, a lot of clients that saw this opportunity would look at contact center and be like, we could get rid of contact center agents and we could build chatbots that would replace the humans. And one, a very, very difficult conversation I'd have to have was to shift their paradigm towards... Sure, you could cost cut, you could reduce your field, you know, your FTEs, your employees in the field, but here's what's happening right now, is that as you begin to use this technology, your funnel is going to get bigger. You're gonna be able to create marketing messaging that talks to more people than you can today. You're gonna be able to create a workflow that operates more efficiently. You're gonna be able to handle more business. Your funnel is gonna get bigger. And so those people... might need to do less things, that's for sure. But you definitely don't want to get rid of them because when your funnel starts to grow, you're going to have cut into the muscle of the organization who would have actually served your scaling growth. And that's something that most CEOs, when I describe it and kind of whiteboard it out with them, they're like, oh my gosh. And most CEOs I talk to now in the Fortune 100, they're not thinking, how do I cut costs? they're thinking, how do I grow revenue? And that's a real sign of maturity for me. Like when someone accepts that this isn't, cost cutting is a temporary tactical measure. And like, you know, we had talked about earlier the concept of like tech, you know, people are laying off teams. I'm like, okay. But it is a sign of a leader who is, or a set of leaders, a leadership team that is being very tactical, very surgical. And sometimes in these very aggressive surgeries, we end up taking more than we need. And we end up hurting more than we need to when a little bit of introspection into the problem. And then most importantly, re-skilling, changing of behavior. How do I change the behavior of my organization so that it acts in a different way? Now, my salespeople, maybe they don't have to spend as much time handling tickets and, you know. talking to clients about problems, but now they can talk about cross-selling and upselling and offering the right product at the right time, even in the right language. Like everything gets bigger and better. And I think this specific organization got very excited about it. They're now, you know, part of the scaled solution when you get to the keyboard step is who do you even use? And so a lot of the organizations have now decided they wanna do this. They wanna amplify their teams and not get rid of them, but re-skill them. But now am I gonna be a Google or a Google house or Microsoft or Amazon? So they're having a lot of fun in my, it's fun for me, but bake-offs to kind of figure out who's the right organization. And the number one question I always ask them first is what their core values are. Because the core values of an organization, if you can align that with the partners that you work with, I've just seen consistent success. And it's a question nobody usually asks. Nobody's like, you know, maybe they, they look at the website and look at the mission statement, maybe they look at documented core values, but they don't really ask like what's important to you. And what's important to you this year is going to be different than what's important to you next year. So you want to create some flexibility in even the partners that you do choose. Because that flexibility is going to give you operational scale, right? You might use a Google model for one thing and Amazon model for something else. Red hat, open shift data science to run it all. Like. Who knows? But flexibility is a key component of organizations that I've worked with that ended up getting to production successfully. Because sometimes we have a leader who's like, I'm a Google person. I was born and raised at Google. I love Google. And that kind of shrinks your opportunity to pick the best and brightest thing to solve your problem. And so I always. I'm a little hesitant when I walk into an organization that's already predetermined that they're all in on a vendor. And I love my vendors, like my children, I love them all. But I also think there's a really strong reason. Each one of them exists, and understanding that is key to scaling success. Absolutely. Well, I think that kind of modularity and future proofing parts of the stack, it pushes everyone to be better when you say, oh, I really like this offering or your pricing doesn't make sense to us anymore. You know, as we scale, like having those conversations and the more and more I build bots and automation, the more it's a human to human, those conversations or those relationships of, you know, these people are awesome to work with. Like, how can I really, you know, their mission, their values align with the way this project is going? I think is, you know, the boats rise, the tides and opportunities to work with awesome people on really cool projects. Yes, and everyone gets motivated when you have that alignment. It creates an opportunity for innovation to scale. Like people, momentum is building and people like hanging out with each other and like working hard. And it's one of the things I noticed during my time at Amazon and even at Microsoft, when we were launching these products, like our teams became like family, like we hung out, we went kayaking together, like it was very familial and you kind of need that. I'm not saying you have to like have a barbecue with, you know, your partners at Microsoft or. IBM, but you should at least feel like they're my people and that you can trust them to give you good information and to ask really good questions on your behalf. One of the things that Jeff Bezos used to say all the time when I was there, he would in his sales kickoffs, he would always say, you know, our job is to innovate on behalf of our customer. Like they can't see where they want to go. They know the problem and they can, you know, forest for the trees, right? They can't innovate because they can't get past. the immediate problems that they have. And that's why it's our job to ask those questions. Like, I know right now you're thinking, you just gotta get through this quarter. So you wanna cut FTEs to do that. But our job is to really show, not tell them, but show them what happens when they build a technology system that can augment and free up 30% of the workforce. And then be solution oriented. Tell them, here's the 15 things that team can now do. Like cross-selling and upselling used to be hard. It used to be very much in the mind of the seller. So I had to train a seller to know how to do that work. Today with these bots that are real time, I could just ask a question, better yet, it can be listening actively to my conversation and dynamically present me, and that's a project I'm working on right now. It can dynamically present me with information and GPT type models, large language models, give me the ability to present it in a super short sentence, right? Old systems, this is not a new solution. We've been trying to solve answering questions on the fly for call center agents for a long time. What's different now though, is that it can say, all right, Noelle, I already know how you like to answer questions. I know the way you talk. So I don't have to, as a agent, I don't have to interpret the information that's coming out of the search engine, right? Now it's gonna already be the way I talk, the way I say it. And it's even a sentence that I could read and it would sound just like Noelle, personalized, hyper-personalized. Like that is a dream. Most people... Again, until they see it, they're like, this is not possible right now. I've had a technologist ask me, they're like, no, for real, have you seen this work? Have you seen it? And I'm like, yes, I've built it a couple times now. So it's such an exciting time because I often use the term bring back the backlog. These are problems a lot of companies have put on the backlog and said, we don't have the tech for that, we don't have people for that, we don't have resources for that. And when you go back to the backlog and you revisit some of these problems, some of them are extremely low hanging fruit in a generative world, right? And so I encourage companies to kind of reinvent who they are and what they're thinking and the problems that they're choosing to go after. Don't presume just because over the last seven years you've not been able to do a thing that it's not doable. You really have to revisit everything that's on your backlog and you might find something that you've long forgotten to dream about doing is not only available, but it's a significant needle mover, right? When it comes to metrics for your business. And that's... I found the smallest things have become extremely large impacts to an organization. Just to wrap up the contact center piece, our metric for success actually shifted. So when I started the project, we were looking at average handle time, which is a typical contact center metric. How long does it take for an agent to handle a call? And our goal was to make that less. But the nuances of that was, wait a second, do they transfer the call? Do they get rid of the call? Do they... hang up, some of us have been there, right? And so instead it was when a call was created and a issue was identified, we were now going to, over the course of two months, we realized we actually want total time to resolution, to resolving it. It's also a metric we measured, it just wasn't the number one metric we were measuring. And this forced us to reduce handle time, but it also forced us to reduce transfers, it also forced us to reduce inadvertent or inverting. disconnections, like all these other numbers had to work together. And of course, we heavily leveraged generative AI to increase knowledge accessibility. So I think one of my always my lessons to people getting into this space is that you don't have to be super technical, but you do have to be data driven, metric driven, because that's how we measure success for an AI system. What needle does it move? What metric does it change? And it has to change it significantly for it to be. useful. We got our PA. That gave us 10, 15% benefit. We're now looking, we want 30%, 80%, 100% gains. And generative ad can do it, but we need to know to be looking at those numbers and measuring them. Mm-hmm. Awesome. Well, and Noelle, I love, or you and I have worked on contact center projects in the past. For people, so some people will know exactly what we're talking about. Like, obviously, obviously. There's some people who are like, call center? Like, wait a second here. And I just want to, for people who might be listening, so contact center, and this is, you know, most of us are end users, right? Like, I have a problem, there's an error on my account, I call my bank, or I text my bank, or I get online, you know, there's a bot on the bank website. And they're actually people. Right, responding. Frequently, these rooms are like, or I've actually, I've never been in one. Noelle, maybe you have, but ginormous warehouse with humans responding to your messages. And they have headphones on and they're responding and they have dashboards. And Noelle and I get to wig out those dashboards, make their user flows better. But I agree with you, I've worked on projects where they're like, we're looking at boost revenue versus cut some of the team. And when... Some people heard they're like, wait, Lisa would lose her job? Wait a minute, wait a minute. Lisa has a key, a top performer. We don't want to lose Lisa. And I was like, does she love her job? Manually copying and pasting codes? Could we upskill her? There's no reason to lose Lisa, as far as I'm concerned. I'm making everyone's lives better. And as you mentioned, data-driven. Let's hit those goals. Let's like. the less time I'm on the phone, like average handling time, like you're telling me you could cut that by 20%, 80%. I got other things to do than fix that thing on my account. Yeah. exactly, exactly. And at the end of the day, you know what is interesting about that is we are now moving into this multi-generational world. Again, a good place for generative technology. But when you're a customer, an organization that has customers, which is everybody, you are going to have these conversations with your customers. And sometimes, depending on your demographic, so I was just over in the middle of the country talking to a farmer, like a grower conglomerate. So a bunch of different growers, farmers, all of them, good 20 years, my senior. And their concern, which I think is a very interesting concept to think about, their concern was like, hey, everyone here is 65 or older, and getting older every year, and approaching retirement, and they've been doing this work their entire careers. So they've been doing it 20 years. So they have this... They have this knowledge of how to do the work that they do. And no one else needs to know, like, it's not written down. It's not documented. This is true in every company, but it was explicitly true. Like when I looked at this audience, it was like, I mean, every year you're going to lose seven to 10% of your workforce that knows what to do. So there's this interesting moment now where we can leverage. And I've been talking about this for long before we got, you know, Models that could do this work. But what happens when I can use a LLM to extract this information? Conversationally from someone who doesn't really like technology. They were all self-proclaimed in my audience. They were like, yeah Yeah, we're not techno. We don't we don't even like technology that much. I Reminded them of course how they use it and how helpful it is. I mean, there's so many cool things We use it in farming in agriculture AI is amazing. There's a whole AI in agricultural project at Microsoft, for those of you who haven't been aware or don't know what's happening, there's so much AI for good. Like all the good stuff is happening, like of course in health care, but also in farming and agriculture, right? How do we do things more responsibly, more sustainably, and even just more humanely, right, in that process. But that all being said, they were, they considered themselves non-technologists and I was like, that's okay. I'm gonna have you just have a conversation and... It would be like you're talking to me, but instead you're gonna talk to this chat bot. Now, lucky for them, most of them have been on Facebook for at least 10 years. So they start chatting with it like a messenger bot, but now I'm asking questions like, oh, what questions would you ask this client? So that I can now feed that same information. It's called expert language models, and there's companies that focus on building these, but I'm taking the expertise of my workforce and building an LLM that can then be used to help newer. younger, maybe earlier in career professionals do the same type of work. Because here's the weird thing that's going to happen, right? As people enter the workforce, these models are going to truncate the amount of time they spend in the field doing hard things and failing, right? Like all the time in the early part of our career, where we're like, well, I guess we're just going to go and try it. And we'd fail in front of a client or build something that broke or, you know, like now we're going to have systems that allow us to build things that maybe they don't. Well, they won't be perfect, but they'll be pretty mediocre. They won't break in the same way that it happened when we built it because it was humans and humans have like a 20% air rate. So, it's just, it's so fascinating because that means that people coming into the workforce now will never ever even organically gain the experience of the people that are leaving the workforce. So it requires us to think very differently about knowledge training, knowledge transfer, where are we gonna get that data and that you have it right now in your company potentially. So how are you gonna extract it, harness it, and then use it as a source of knowledge for people? And I think over the next two to five years, that's gonna be a big area where it's, again, I'm not touching a specific business problem other than maintaining competitive advantage through the industry expertise I've developed in maintaining employees for 25 years. Like if you have an employee that's been there 25 years, start asking questions. Write down everything you hear. because they're a gold mine and they will eventually leave. And I think over the course of the next few years, we're gonna see that pain. I saw a presentation and the guy was like snapping his fingers like this. Every time he snapped his fingers, he's like, that's someone over 65 choosing to leave the workforce. And I was like, oh. And so it's really, it's not a problem we think about, but it's a problem uniquely good to be solved with a generative platform. So something I thought that was just interesting way of looking at things. That's amazing. Well and Noelle, I don't think you know this, but I'm working on an agriculture project right now with a someone who's also significantly older than me. His zone of genius is in agriculture. It's in farming, which I have very little domain knowledge of, but he needs help and today's a lot of pen and paper. Today, it's so much manual stuff, but the scale operations, he knows he needs help. People frequently are asking him, what's your AI plan? He's like, Who else is going to do that? Because it's not going to be me. But the deep domain expertise he has, and how do we document that? How do we share that? It's a brave new world. I think, honestly, he's a fabulous customer because the mindset. It's not like, uh-oh, Joan, you're scaring me. It's like, ooh, let's go do that. I want to learn more. But wow, I hadn't thought about documenting right exactly the huge knowledge base that they're just holding. Yeah. and it's already semantic. You know, like it's already the semantic layer. It's, we haven't, we're just scratching the surface but I think anyone who's moving into this space, A, number one, you could represent that expertise and be part of a team that harnesses it within an organization or even, I mean, my dream would be we don't do it inside companies that will silo it off and protect it, that we do it in consortiums and we do it in industry, academia partnerships. That's like. a utopia. But most organizations instead are going to be very closed off and be like, hey, we have the best agricultural expertise, so we're going to create an LLM and then we're going to sell the LLM. And we lose a lot there because a lot of, I always tell people like data, modeling is not your secret sauce. Building an algorithm, that's not your core competency. Your specific data, your understanding of the market, yes. we should do a lot better job of pulling together what we know about how this data is structured, how it works, how it connects to each other. And right now, I sit in a company, if I'm at Accenture, I know that there's a team at Pric PWC, that's doing the exact same thing, and I'm not allowed to talk to them or share knowledge with them. Or, in my mind, and again, I think you and I share this, I just have a heart for us working as a humanity. to make humanity better. And so to watch these silos and echo chambers develop, I'm like, can we just go outside and hang out in the park and talk about what we're doing? But so that's organizational and that's decades and decades old, right? That's systemic, it's very hard to fight, but there are these opportunities. So I love, you know, like this, the idea of this podcast, right? Bringing people and giving them a chance to share what's working for them. to give us as listeners a chance to like see, you know, on the other side, see through that curtain, if you will. So it's super, super helpful, but it's a challenge. It's a challenge I face every day. Yeah, no, it totally is and I think the irony is that people keep swapping companies, going to different jobs. So there are stakeholders that have that knowledge across the base, but certainly not. Well, okay, we've looked into some projects. How did you get into this field, Noelle? I think I know a little bit, but I'd love if people are curious. Absolutely. So, um, here's the, the truth of it is about 10 years ago, I was at Amazon. I was a principal cloud architect. so nowhere near AI, uh, and Jeff Bezos sent out an email to everyone and said, Hey, we're starting a new project. And it was like a pet project of his. It wasn't rockets, right? But at this time it was this little box. It hadn't even been launched. There was no prototype. Like, I think we had a couple of prototypes, but it was a box. It had a curtain over it. And they were like, we want to create the Star Trek computer. Um, and at that time there was no VPs. There was just one director and he was hiring people. And I think I was like the eighth or ninth person, uh, to join that team. I then became one of the early members of the skills team to start building applications. Um, but the main, I always tell people the only reason I went and did that thing was because, a, I love Star Trek. So Jeff Bezos did a good marketing pitch there. But I also had, you know, you know this about like, in my personal life, I have a son with Down syndrome. I now have a dad that I take care of who's aging and has traumatic brain injury. So he, you know, we have heavy needs for automation inside our home. And I looked at the Alexa project 10 years ago and I was like, oh my gosh, if that worked, that would be really cool. And we have now over 130 devices in our house. We are voice everything. It's- really impressive. One day I will like do some kind of like e-Hollywood story on like all the devices we've bought, failed, that work, that don't work. But that journey started with literally an alignment of what I cared about, my son, my dad taking care of people, and a dream I had, which is around science fiction really, like Star Trek, Asimov, iRobot. Like I just cared about this rather than doing what many of my peers did, I said, yes, like many of my peers. And I was like, hey, we should go do this together. And they'd be like, oh, oh no, I just got promoted. So I'm gonna just hang out where I'm at. And I was like, wait, this is exciting. So they let me come and play and do that. But I went all in and I do this in every role. I've had three or four different curves that I've been part of in my career. I was part of the early web services, like initiative that happened at IBM a decade before that. I was part of big data and Pivotal when that happened. So I'm always kind of in this moment when people are trying to do a new thing. I tend to be like attracted to those types of companies. And so Amazon Alexa was doing this new thing and I jumped right in. And here's the cool thing. I didn't know anything about any of that world, but what I did have was in a extreme desire to learn. Um, but more importantly to demonstrate. And luckily for me, this was a application development team, right? So I got to build apps and that was the best way to demonstrate my expertise. I always tell people, sometimes you're going to be a UX designer. Sometimes you're going to be, you know, project manager. Sometimes you'll own the product as a product owner. In either case, use it every moment as an opportunity to teach people. And I just did this with generative AI. So I just got a chance to grow a generative AI practice at Accenture. And the very first thing I did, I mean, I was three weeks into the company and I sent out a note to people I do not know and said, Hey, if anyone wants to learn what I know about generative AI, because I was already the, I was on the beta of GitHub copilot during COVID. So I had like two years of experience and everyone was like, what is this new thing? So I was like, I will teach you from eight to 10 every morning for two weeks, what I know, and we're going to build the foundation of a new practice and we're going to all be very busy this year. And all of that ended up coming true. But the key was, did I know how to do any of that? No, I learned by doing. I went to Microsoft and said, what do you have? They gave me three GitHub repos. Now there's like dozens, but I got three GitHub repos and I just started building. So it's a bit of a mind shift from, hey, will I go to a bookstore and read a book and then wait for a manager to come to me and offer me an opportunity? I mean, I became a hunter for innovative projects to become part of. And... I got that opportunity at Microsoft, but I kept my reception open. I constantly was on the lookout for, I have tons of Google alerts on AI for good, AI for humanitarian efforts, AI for agriculture, all these things that I care about, AI for accessibility. And as projects would come up, I would do something, which most people apparently don't do, I would reach out to them on LinkedIn and be like, hey, you're doing cool stuff, can we be friends? And that friendship, over, I mean, Lon Guan, who's now the chief AI officer at Accenture, she knew me because we were befriended years before in one project that I was doing. And I saw she was doing something similar. And I just was like, hey, we should be friends. And that's it. And then later she was like, oh my gosh, you should come run this team for me. So really interesting approach, but I always encourage people like none of that was because I was classically trained in AI because I'm not. None of it's because I went to Stanford or a big university. I didn't. Like all of it was because I had... purpose that was driving me to be very, very active in learning. Um, and that I also was really excited about the problems I was choosing to solve. Oh, right. And the keyword I was choosing to solve them. I always like to let people know you do have a bit of agency here. I know it's a rough time and a lot of people are looking for work, but you do have a choice and the better you can articulate what you want to do, the better a company can find a place for you in their organization and clarity of thought. is a special super skill that not everybody has. So I encourage you to focus on harnessing that part of your brain and your emotional intelligence because I feel like people who have emotional intelligence and can build AI solutions, they're few and far between but we need them more and more than ever. Yeah, oh, Noelle, that was fantastic. So just for those listening, I heard massive hustle, the ability to, I think, just have the courage to put it out there, which some people really struggle with, bang down the doors or at least send a DM, hello, we'd love to be friendly. Yeah, exactly. And I think also that like honing down or at least having a niche or saying, this is the one unit of thing I wanna do next. Because as you mentioned, what is this year versus next year, let's find out. But I think I know a lot of people who, and myself included, there's so many distractions, right? There's so many different things. But just defining what is the next, and finding stakeholders and friends who are like, let's do it, like, person A to Amazon Alexa. So freaking cool, Noelle. OK, and also a question I get very often is about different certifications, different You mentioned Stanford. Do I need a degree? What are building blocks towards getting into these rooms? Yes. So it does depend. So when I was joining Amazon Alexa, I was a, it was a startup, right? So nobody needed anything. They were just looking for hungry entrepreneurial people. And I definitely fit that mold. As we get more mature in the space, I personally then decided if there was a certification that I could get, like a industry certification, Microsoft, Google, Amazon, I would go after it. So I became Amazon certified solution architect. I, you know, got their cloud became an MVP at Microsoft. Like how can you, if you don't go to a Stanford or an MIT or even any university and get a stamp from a third party saying you went through this work, because it is just an aptitude test. Like people wanna know, do you have the aptitude to build something awesome? And school is a great way to check for that. But there's also lots of other ways. And so going through Microsoft Learn courses, going through LinkedIn learning courses, now I like to build them. But back in the day, I used to just, I'd knock them out. And here's, I would say, I think there's a statistic out there from LinkedIn, but like 8% of people actually finish things they start on LinkedIn or on any of these learning platforms. Same is true for books, less than that I think now, but less than 10% finish books that they start. With all the intention, I'm a book finisher, so I like took that as a challenge, so now I finish books I start, to my detriment where I'm like, oh my gosh, do I have to finish this? But that being said, you have to be a finisher and you have to know, Like that's what people will value. So if you want to go to school, great, but go there with the commitment that you're gonna finish it. If you're going to, you know, if you know you're not great at school, I don't like school. I don't like learning in a classroom. I like learning with my fingers on a keyboard. So I'm much better at like bootcamp type things. Or I spent three days at Stanford with their foundation model team. That was hugely valuable to me. I learned, I think more than most people because I'm a sponge in that environment. Other people want to sit and listen to lectures because they take a lot of notes and that's how they acquire information. There's a book called Thinking Fast and Thinking Slow. Like everyone thinks different and it's okay wherever you are, but the more you know about how you learn, codify that knowledge with some third party saying, yep, Noelle knows that. And it's not even, I like it. I like the idea. It's confidence building for me every time I pass a certification, but I am a certification person, so I'm a bit biased. So I always say get one if you're in this relatively new or even if you've been in a while and you don't have them Getting a certification from every platform like the large players Microsoft Amazon and Google. They all have introductory digital leadership type certification so AWS has I think it's called cloud practitioner Azure has Azure fundamentals. They even have an AI fundamentals, which of course I'm a big fan of and then Google has a digital leadership certification. None of these are like deeply technical. You don't have to code live in front of a human. Like there's no anxiety. They train you well, but it gives you a vernacular and a stamp that becomes very valuable when you're starting to have conversations either as an employee or even pitching to organizations, right? Where you want to get funding as a startup, like having the understanding of how companies talk about this thing. Because of course they all talk a little bit differently, but then you'll realize after getting three that they have a lot of the same core. like same core understanding, same core learnings. So that's, I am a big fan. I say go for it, especially if you're not one to go back to school. I mean, if you're a school person and academic, like do that. If you win there, you should do that. I think Joan is very much like that. I love it, but not everybody's built that way, but if you are, you should do that. Like there is no, no one way is better than the other, but it does benefit you to be kind of quantified. by someone other than yourself. It's better, and that's just a true marketing, right? In marketing, it's better for someone else to say you're awesome than for you to say you're awesome. So I'm a big believer in that. yeah, I certainly I got, I got the masters and then I got the PhD because I wanted to be in R&D. And so many of the opportunities I saw, like the first line is PhD in this is like the first one on job req which is actually not how the world works these days now that I'm learning about how jobs actually get gotten. But I want I want to be considered for those rooms, I want to get those offers. And certainly, it has opened tons of doors for me. Not that everyone needs to get a PhD, there are many ways. You also mentioned, was it some place you got three days on site that you learned a bunch? Was that it? Stanford has a program, MIT has a similar one, and I love them both. I like the Stanford one because they have an organization called Human Centered AI. From that, there's an organization inside of their mini organization, all around foundation models. So they're the organization, the team of researchers that coined that phrase. And Percy Liang is kind of leading that effort. I enjoyed it because as a practitioner, I'm just kind of like putting, you know, I always say like rubber to the road, like I'm moving code into production. But going there, I got to see and hear why and how things work the way they do. And a lot of the research that goes behind like things that show up in product, if they're in my mind, if they're really good, often came from a research background. Like we proved that it worked in this way. We now are productizing that for the rest of the world. Very few times we just like spool it up out of nowhere, especially in AI. lot many of the models that people are like this is new and we're like well it's technically based on a decade long you know research project that just completed at Stanford so I love that program it's called the um for those who are taking notes it's called the Stanford Foundation Model Scholars Program so you can reach out to um HAI which is the human centered AI organization there and ask them about it they open up a couple cohorts a year But it's fascinating. But again, there's equally great academics at MIT that do a similar type of foundational workshop. There's also, of course, and these are lower dollar because they're meant for a broader audience. So you're not going to get super deep technically. Like it's not for researchers. It's for people like executives like me. Harvard also has a program that's available, but it's like six weeks. I'm a, you know, Three days, I'll work like 40 hours in three days. But six weeks is a long time for me. So I think the industry, the educational industry is really starting to change, right? Remember, there was only a few years ago, Harvard only did a certain type of education, right? edX, I think, was what launched us into this new world of like accessible education in many different modalities for people. So be aware that if you look for it, edX has... A whole bunch, I think 20, 30 Ivy League organizations, colleges and universities have contributed free content there. So that's where I go to learn. It's not where I go to get certified, but it's where I go to learn. That's awesome. Well, I'm holding these in the show notes. Thank you for the names. I'll help us find that. Well, coming to the end of this, what advice or steps would you give someone, depending, regardless of where they're coming in from the skills that they have, what advice might you give? Yeah, I think one of the principles that I lean into heavily in the projects that I do and when I bring new people on my team is really cultivating this idea. I actually learned it in Amazon as one of their leadership principles, and it's learn and be curious, but I take it one step further and I've reframed that to learning by doing. And I learned to build applications for Alexa. Not because anyone knew how to do it, because it's just like generative AI was last year. No one knew how to do it yet. So I had to go and look at people who knew how to do it a little bit, take what they did and learn as quickly as possible, but then let my own mind and ideas pull these GitHub repos together and look at patterns and kind of tinker. I think one of the, it's very generational I find, but tinkering, being able to play and experiment, being able to harness your inner researcher. These are really incredibly important skills that you won't put on your LinkedIn profile, but that will give you the ability to communicate what you know, how you know it. Like projects, I did hundreds, hundreds of projects on GitHub and they weren't mine. I went and took someone else's project and then I rebuilt it. And I literally, line by line, some of the code I didn't even understand because it was in a language I hadn't learned yet. And I was like, oh, Node.js. What is this? And I just started writing. I'm like, ooh, there's a lot of similarity with Java, which I know really well. Not the same, you know, obviously technically, but I was able to kind of pick it up as I went. And I think that's really important for people not to realize, like you don't have to like go into a class and then wait for the class to be over to like go on this traditional path. It gets extremely accessible. Anyone can start learning. There's lots of open communities. Speaking of communities, I have a community called iHeartAI. I wear a shirt that says I Heart AI on every stage. And so I have a community, I'll give you the URL so you can I think the other thing I say is like, get into a community that's not directly tied to your profession, because there's a bit of psychological safety in a world where people aren't like measuring your success for performance purposes, right? And you need that. today in the world of Fun story, I was at Microsoft Ignite and they had a getting started in AI booth and no one went to it because no one wanted to be, this was my own personal interpretation, nobody wanted to self identify as a getting started in AI person. However, I was right next to it in model evaluation, which is one of my passions, and they were asking me all these getting started questions. And it just made me realize that we don't have psychological safety sometimes in these companies that we work for in the industry, you know, echo chambers that we're in. So join a community that will celebrate you asking questions, even if they're silly or you think everyone knows it. I mean, there's so much to be learned. And if you don't find a home where you feel like you can just learn along with everyone else and find people that are kind of kindred spirits with you. I feel like it's just a much harder role. You could still be successful. I know lots of people who are, but it's way more fun when you find friends that you can hang out with. Joan and I, we've known each other for a long time and we've found these groups, right? We've sat in these rooms together. I think it's just so important. So yeah, so iHeartAI, it's on school. I'll send you a URL, so anybody who wants to join. I'll be doing five days of fun content, just teaching people what I think, just like we covered here. in a little bit more depth because we have five hours to go through it. So yeah, but I'm a big fan of that. Like just get with people you know, like, and trust and learn together. And that is a really critical, I think, opportunity for you to shift the trajectory of your career or business moving forward. Yeah, that's beautiful. I dig it. Well, Noelle, if people want to find you, where can they plug away? Where can they find you? Yes, I always encourage people to connect with me on LinkedIn and not follow, though you can follow me, but it hit the little three dots and actually connect so we can message back and forth if you have a question or if you see I'm going to be in a town. Many, many times I'm in a town and someone I've been friends with on LinkedIn for a couple of years will be like, oh my gosh, you're in my town and I'll have, you know, we'll go and have Starbucks together. So would love to be connected. And the other thing you could just find my name, you could just Google me probably, but noellerussell.ai. It's where all my things are. So my community, my newsletter. I launched a new podcast. So I'm super excited to be on this one called Good Morning AI. I have a podcast we launched last year called The Lamplighter Effect, which is a leadership podcast. So there's lots of stuff. If you enjoyed this time, you know, definitely I welcome you into my family. And it's been truly a pleasure talking to you. Well, thank you so much for coming on this podcast. Noelle, it's great to see you. It's great to learn from you. Thank you so much for your time. Yes, thank you. Very grateful to be here. Have a great day. Oh gosh, was that fun. Did you enjoy that episode as much as I did? Well, now be sure to check out our show notes for this episode that has tons of links and resources and our guest bio, etc. Go check it out. If you're ready to dive in to personalize your AI journey, download the free Your AI Roadmap workbook at yourairoadmap .com / workbook. Well, maybe you work at a company and you're like, hey, we want to grow in data and AI and I'd love to work with you. Please schedule an intro and sync with me at Clarity AI at hireclarity .ai. We'd love to talk to you about it. My team builds custom AI solutions, digital twins, optimizations, data, fun stuff for small and medium sized businesses. Our price points start at five, six, seven, eight figures, depends on your needs, depending on your time scales, et cetera. If you liked the podcast, please support us. Can you please rate, review, subscribe, send it to your friend, DM your boss, follow wherever you get your podcasts. I certainly learned something new and I hope you did too. Next episode drops soon. Can't wait to hear another amazing expert building in AI. Talk to you soon!