Skills@work 2025: Impact Accelerator #2 - Human-Driven, AI-Supported: The Future of Skills Activation

So as, as a quick intro, and first, let me, uh, put up our presentation here, uh, if everybody can see that. So, let me just start with an introduction. Uh, my name is Matt Mogue. I'm the founder and CEO of Career Bird. Uh, I, uh, based outta Chicago. Uh, this is the fourth company that I've, uh, founded, uh, over, uh, the last, uh, years. Uh, the other, uh, three companies all, uh, grew from an idea to have hundreds of employees. And I had the, the unique opportunity of rebuilding our, uh, people and culture teams each time, worked with incredible leaders and, uh, really tried to focus on not just talent acquisition, not just people operations, but executing a really great talent strategy, uh, that would make sure that we had the skills that we needed. In some cases, the companies were, we were, we were doubling and tripling from one year, uh, to the next. So it was always a, a tricky, uh, challenge. And one of the things I learned is that, uh, HR is chronically underfunded and under resourced, uh, and doesn't always have the tools, uh, needed, uh, to get some of the jobs done. Uh, and when it came to talent development, specifically, what I saw as CEO being a software person is, is that, uh, there, the mid-market was really being ignored. So software solutions that were specifically tuned to be both affordable and accessible and easy to use for companies with hundreds or thousands of employees, not with, with necessarily with tens of thousands. Uh, so we set out in the age of AI to build a new talent development platform that specifically meets the needs of the mid-market. And let me get into what I see as some of the larger pictures, and many of these things have been mentioned. I think it's really helpful here, and this presentation will be sent out. But the, some of the, the statistics behind this are, as I think all of us have seen, there's a, a giant generational, uh, shift happening. And expectations are different among millennial and Gen Z as you can see here. Uh, career progression and skills development ranked number one among the expectations of the workforce. It is the number one reason that people choose to leave organizations, uh, to put in a more common way when employees are asking, when do I get promoted? How do I get promoted? And what are you doing to invest in my, my growth? Uh, many organ, many mid-market organizations don't always have a really good, clear answers to that. And that's part of what leads to people leaving at the same time that this major shift is happening. Obviously, distributed work, uh, people working, as you can see here, of companies are hybrid. There's only of companies that are now fully in person. That's really led to a dramatic, uh, shift in how managers need to manage, uh, and have put more onus on them for understanding exactly what needs to get done, uh, and what skills are needed to get that work done. Uh, during the same time, the, the, the significant changes really over just the last four years to the number of people who are in the fractional and gig workforce is significant. I, I, I was shocked to learn the other day when I talked to somebody from DoorDash, believe it or not, DoorDash employs million, I'm sorry, million people who deliver food and Uber employs million. Uh, trying to think about this, the, the scale of those kinds of numbers. And in the midst of all of that, there's just this incredible pressure on every dimension, technology, regulatory, customer experience, competitive pressures for businesses to react quicker to change. And I, I think, has having been a CEO for the last, uh, years that it is, uh, it is a bygone era where HR is more tactical and just takes care of, of payroll and benefits, and HR really needs, um, and deserves a seat at the table with the executive team. And in just the last week in this Wall Street Journal article links to this, uh, you have CEOs of the largest companies in the United States, uh, talking about how, and this is a direct quote, A AI is going to change literally every job. That's not a vendor. It's not a procrastinator. It, it, it's not, uh, somebody deep into this world. It is a CEO of a major company employing hundreds of thousands of employees who sees the profound change that is happening. So with that as a, uh, the, the bridge from those macro changes to the reality that the skills that employees need to do their jobs are changing so significantly, and from this very recent, uh, future of jobs report from the World Economic Forum, of key skills required in the job market will change by . And this is, this is a, a direct quote from them as well about the need for continuous learning and upskilling and reskilling, all things that we have talked about today. And the, the, the, one of the questions here is, what's the link between an Agile organization and a skills first organization? And what does the skills first, first organization really do? And to me, it's about connection, uh, between, and, and this is where I think sometimes gets missed in conversations, that you have to think about how you have designed your roles and the training and the career mobility. If you don't connect those things in a very, uh, thoughtful, intentional way, they're much harder to get traction and to move forward with. And that's part of what we've tried to, to build into the new platform that we've created. So the steps that I, I would see as an org for organizations, uh, moving toward becoming skills First and Agile is number one. And this has been talked about several times, focus on job design and job architecture. But I, I do wanna . important thing out is that, uh, a lot of people will talk about job architecture and families and job tracks and, uh, leveling and all those kinds of things, competency models. What they're often missing though, is the, is the fundamental design of the job description itself. What's on it? How comprehensive is it? Who's doing it? How often is it updated? And one of the things that that spurred me to start this business was I was part of an organization that went to do a, a compensation and, uh, pay equity study. And we realized that more than half of all of our roles in the organization did not even have a job description. And those that did, they were mostly completely out of date. And when you go back to that previous slide about the, the, the, the rate at which people, uh, uh, skills are changing, that's really significant. But step one, you, you're, you're mapping job design and job architecture to a continuous learning program where you're mapping the skills that you need to learning resources. This is another area where I think ai, uh, is gonna bring whole new, uh, opportunities rather than do generic training around a horizontal, uh, competencies or manager training, et cetera. Uh, uh, professionals in this space now have an opportunity to train on domain specific skills that are specific to a certain role and a certain level, and to source those learning resources from across the web on a real time, uh, basis. We talked a little bit about the skills first orientation and how it's, it's really about connecting different functions there, that this, and this was mentioned earlier as well, about this notion of blended talent models, of focusing on internal mobility. Uh, and, uh, considering that when you really define the skills needed for your roles, that that then allows you to tap into a geographically distributed pool. And I think something very, um, it's both, uh, very notable. And also what though, unremarkable in some ways, I guess, is just on this webinar alone. There are, uh, presenters and attendees from absolutely all over the world. Uh, and the fact that we're all coming together around a common topic, I think is, is really telling of where the world is going. So the, the action, the very practical action plan that I would recommend for you to assess your own skill readiness is, one, do you have a process to determine what skills are required? And do you have a way to automate that? Because when you scale it across hundreds or thousands of roles, it can become very labor intensive, and you need a way to both del uh, a workflow platform to delegate that and make sure that you have approval rights and governance over that. Number two, once you've determine what skills are required, what's the method that you're gonna use to determine the proficiency of the employees against those required skills? And, and once you've identified, uh, what that proficiency is, can you take action against it in terms of, uh, learning and training? And can you have an expansive view of what learning and is not just a course, it could be a podcast or an article, or an applied learning opportunity. And of course, the fourth one here is to motivate employees and fully engage them. Can you demonstrate them to them how the, the definition of their very job and the skills they have for it, and the skills that they gain through learning are then connected to career progress. So I'm, I wanna give you some visualizations here, uh, that, that will help bring some of this to life. So when we talk about job design and job architecture, you'll notice on the left side of this slide that we're, you're looking at, have you defined responsibilities, technical skills, behavioral competencies, leadership competencies, success metrics, education and certification, all of those kinds of things. What you'll see in the vast majority of cases is that even sophisticated companies don't do a good job of determining the skills and competencies required for a role, and particularly not the technical skills. Uh, that's o often left off of both job postings and most internal job descriptions. And if you don't have that, it's very difficult to have a skills first strategy. Um, the other part of this too is you'll notice under job architecture of all the common things that you would expect, uh, around leveling and job tracks and job families. But do you have a, uh, platform of record that allows you to track all of your changes and also, uh, a workflow and approval tool, uh, so that you can delegate this responsibility out to your various business owners? The next question around assessment and gathering proficiency levels is both, can you ask both the employee and the manager to weigh in on what their level of proficiency is with all of the skills that are on their job description? And allow the manager and the employee to calibrate around that and really understand if there are differences between their view. What are the reasons for those differences that then we, we talked about this idea of mapping specific technical skills and competencies on a, on a role by role basis to a range of learning opportunities and resources. And very importantly, enabling employees to share and curate those learning resources. The, the world is moving simply too fast, uh, to rely on outdated and static learning resources that were created years ago that are no longer relevant in today's, uh, AI disrupted world. Uh, and then ultimately, do you have the insights for strategic workforce, uh, planning? And is this something that both, uh, HR can see as well as managers at a team level, uh, and even employees can see about themselves that leverage the skills database that you've built as a result of defining your different roles? Uh, and what ties all of this together is, is taking your skills that you've designed, uh, and, and, and mapping them to the career path for the individual so that whether it's within their department, within the level that they operate in, uh, or within the job family or job track, they can start to map and see exactly side by side each job and, uh, how it compares and what are the steps that they need to take in order to improve in that area. So where I would land all this is where we started about the, the importance of HR taking a seat at the table, being a strategic voice in the larger strategy of the organization. And when you think about the larger efforts of the CEO and the C-suite and the other leadership teams where they're mapping together first, their, their overall strategy and strategic vision, putting together usually a strategic plan on an annual basis where a HR should really be coming to the table is, is mapping workforce development needs to that strategic plan. Uh, and with all of the things that we just talked about and understanding what skills do we have? What skills do we need? What do we need to hire for, what do we need to train for? Where could ai, uh, help us, uh, fill some of those gaps? And what, where are there fractional and gig related workers that are, that are opportunity for us? But at, at the end of the day, uh, this is where, uh, a HR really has an opportunity to, to step forward. So I, I think, uh, Zach, we're, we're pretty much on time, and I don't know if we have time for questions or not, but that, um, uh, appreciate the opportunity to share some of that. Yeah. Give it up for Matt. I mean, I appreciate you breaking this down. And it's also just reassuring hearing a CEO who noticed this at other companies that you scaled in the past to the point that you saw so much value in it, that you built a new company around this. Right. Um, and I would love to ask a follow up question specifically, you know, with your point of view as a CEO, we have a lot of HR leaders in the room that maybe don't have that seat at the table, or they haven't built maybe that strategic relationship with the CEO at this level where they're struggling to kind of get that CEO bought in. And I'm curious of how you would encourage some of these leaders to approach C-suite or approach CEOs to say, you know what, yeah, you're right. This is really critical to the strategic plan that we're, we're rolling out for next year, and we need to make sure we have a talent strategy that aligns with that. How do, how would you encourage some of these HR leaders to kind of surface this conversation? Yeah. Or get the CEO really bought in? Yeah, it's a good question. There, there are three things that I would highlight that every CEO will respond to. And one is, uh, do you have insights or ideas about ways to help the company grow? And how do you map the talent and the skills to those growth initiatives? And very specifically, uh, draw that, that connection. The second is, uh, can you identify as ways for us to save money? Uh, right? So if you can grow and you can save money, you're, you're obviously helping to contribute, uh, to that as, as well. And it's, and being very outcome, uh, driven, uh, in your, uh, presentation of this, that, that's to me why, uh, having the skills database, uh, and mapping it to roles and understanding that, and the, one of the biggest topics on the minds of nearly every CEO in every industry, and I think it's only gonna get more significant, is how can you help drive the conversation around ai? You, you don't need to be the CTO to do this. You are, you are the chief talent officer, uh, and your, your, your job is to figure out what skills, uh, do they need. And if you can break those AI challenges down at a departmental level, what does marketing need, how does marketing need to use it? How does, what does sales, who within marketing and sales is gonna lead the charge? And have we made sure to update their job description with the appropriate skills? And can we use all of that insight to then identify, uh, roles that could potentially be augmented, uh, with ai. All all of those things I think would be very well received. Yeah. Well, this was incredible, Matt, where can people connect with you afterwards? One, I'll share your LinkedIn, and I know you were generous enough to, we talked a lot about like pilots today and how we can build momentum around those things. So can you share a little bit about ways that you're supporting people kind of get started in this path? Yeah, for, for sure. And the, the platform, we, we've built job architectures in hours for companies with thousands of people. We've inventoried all their skills, et cetera. So what I told Zach is that we're willing for the first qualifying companies from here to do a day free trial with you and just show you how we can help you build that database. If that's sort of starting as too much, we can also do a skills audit for any individual department in your organization. And even more if you just wanna see one small piece of it, uh, send us the, the, the title for any role in your organization. And we'll send you back a link to a comprehensive job description that's with skills and competencies, leadership qualities, responsibilities, so you can get a sense for the kind of depth and richness, uh, that you can generate, uh, to define the roles in your organization. Uh, you can reach me at Matt Moog, M-A-T-T-M-O-O-G, at career bird.ai, or you can also just go to our website, career bird.aicontact and fill out the form there. I mean, for those of you that have already started trying to do skills audits and you have crazy spreadsheets, I've done this before, it's nearly impossible to scale across your company, and it's just a tedious act. So one, I mean, just use this as like a step forward and get a free skills audit, like help kind of get some momentum going in the right direction. And then, as we kind of talked a lot about today, getting pilots and small wins within specific use cases is how you start to really get buy-in as well with the C-suite. Like, okay, if this is starting to kind of as you shared save costs or improve productivity for a certain team, we better believe they're gonna start looking at scaling that across the company. So Matt, thank you so much. This was awesome. I appreciate you being part of the network, and thank you for the time today. Thank you, Zach. All right, everyone, let's keep this moving. How about that? I mean, so cool to be able to start thinking about what this actually looks like in practical sense and how do we start to scale that and leverage techno technology to do that. So next up, and Taylor, I already see you up, so let me bring you up. My man. Appreciate you being here with us. If you, if you don't know Taylor, he was already with us a few months ago, maybe at this point, talking about how AI and some of these things are impacting our skills strategy, our workforce strategy and things like that. So really excited to bring 'em back to share some more tactical insights. So Taylor, thanks for being here, and I'll pass it over to you. Thank you for having me. I am on vacation, so if you do hear some kids in the background, that's what we got going on here. But, uh, Zach, excited to be here. Thank you for you and, and chief for having me back. I'll go ahead and share my screen and then of course, ask the question that we always ask at the start of every Zoom meeting is, can you see my screen and let me get it put here in present mode. Is that showing up for you? Someone? Anyone? Yes, I think we're, yep. Okay, perfect. Uh, fantastic. So, uh, thank you for again, uh, for having me here. Uh, my name's Taylor Bradley, the head of the OR talent strategy and Success at Touring, and we are an AI infrastructure company. So of course, when Zach reached out, we are gonna talk a little bit about ai. And what I wanted to highlight was just three practical takeaways. I'll try to get through my slides, uh, very briefly so that we can open it up to questions. I'm here for you all, so if you have questions in this topic, even spicy questions, looking forward to those. All right, we'll go ahead and click here on the next slide. So I'm a member of CNBC's Workforce Executive Council as well, which affords me the unique opportunity to attend a CEO summit each year, which is made up of CEOs from Fortune companies and Russell , some of the world's leading CEOs. And, and this year's summit, just a few months ago, these were still some of the common questions that I was getting. And more often than not, when I meet with executive leaders, like many of you on this call, uh, with some of the client work that we do, the most common conversation that I'm still having is we're absolutely in on AI being an important tool in the toolkit, but we're still looking on how do we scale this? How do we build proficiency in the organization? So I wanted to take just three nuggets out of those conversations that I have, share them with you all, and then of course, get to your questions. Those are the AI maturity model, a self-assessment that we share with our folks, how jobs will evolve, so that helps inform how you actually implement AI use cases. And then of course, our idea methodology that we use internal at Turing. So the first off is the maturity model. Many of you are l and D leaders on the call. So a good maturity model is exciting for everybody, including myself. We have three dimensions that we share leadership alignment. So top down workforce engagement, bottom up. And then of course, AI integration into your workflow is existing today. This self-assessment is super easy. There's no right or wrong answers, but it helps inform us when we're working with our clients or if I'm going out and just chatting with one of you over coffee of where is your team or your department or your organization today. And then we can tailor a plan to help scale AI within your organization. And here is what the self-assessment looks like. Nothing too unruly, uh, you go through should take you a minute or so Many of you are probably already selecting your level as I speak. And again, there's no right or wrong answer. And we've seen all different matrix here of what people will place themselves in. I think one key thing to acknowledge for an organization is that each department, and even within the department, each of your teams may be at different levels. And so that is important to have that awareness so you can start building out what skills need to be improved and what support you need to provide to your organization. Then of course, we get into the question that I get asked most often is how jobs going to evolve? So I ended up making this little infographic and prior to we'll use chat GPT as a, a, B, c, a D moment in the world of ai, 'cause AI has been around for decades, but generative AI in this format that's so consumer friendly, really did disrupt all of our lives. And so prior to that, a human job, what is a job? A job is made up a collection of tasks and responsibilities. Uh, we've seen a lot of those in the form of job descriptions or as Matt highlighted, many of us probably don't have job descriptions for every role and we should get on that. Uh, so connect with Matt on that point. But that is what we like to understand. First is, in a particular job, an instructional designer, a people operations specialist, what are the responsibilities and tasks? We then work with my team. I've worked with our people operations team to map out their role and ask the question, what parts of their, their, uh, their role would they desire to get support from ai? And there's two mechanisms you can get support from AI primarily is augmentation. So human-centered AI assist, this is the most common support of AI or generative AI that we are using today. And then of course, automation. And this is where it's machine centered. AI is acting independently with the goal of understanding from the human in the role, which of their tasks would they like to convert from a gray human-centric dot to a blue either AI, augmented or AI automated role. Understanding that desirability of which parts of their job also, uh, plays into a positive change management experience because you have a insight into what would enrich this person's role, what jobs can they offload to ai so they have more time to focus on more complex or human task in their role. And that's exactly what we did with our people operations specialists. Then I get asked a lot, what's the long-term vision look like? Is it going to be Skynet or something along those lines? What I will say as a preface to this is pending any watershed moment in quantum computing or artificial intelligence that we may not be aware of or they could come that could disrupt all of this. But let's assume that we're on the trajectory now. At some point in the future, there is still going to be a human in the loop, a human in the center, but some of our managers, some of you may start to end up managing more AI agents, uh, on your team than human professionals or have augmentation from AI assistance. I draw a differentiation between assistance and agents. 'cause there is a very technical term for an agent. An agent is often wildly overused, uh, in today's context. So that's why I tend to say you may have more actually AI assistance helping you with your augmentation rather than a a AI agent, which is making independent decisions and forecasting, uh, and a bit more nuance there. So we continue on right now. Claude did a fantastic research. They have an amazing research team, uh, and I encourage you all to look up their report. Uh, of occupations that they've identified are at least using of ai, either augmentation or automation for their associated tasks. So of these gray dots are being done by AI today in about, uh, three of the roles I do get asked how much augmentation automation. So augmentation and about automation is what we're seeing right now in the successful use cases of ai. Then we get to the third and final, and then of course we'll get over to some questions here from you all is this is what we train all of our team members at Turing. And on the people team especially, is how do they develop an idea for an AI use case? And that's identified design, experiment and action and AI use case. So that's turning one of those gray dots into a blue dot. And so what we do is we make sure that they have the right AI literacy to identify and pinpoint which job tasks are suitable for ai, either augmentation or automation, and then design. We do have, uh, some great enterprise licenses, which provides a safe environment for us to work with our data. Uh, both Google Gemini and uh, chat GPT are available to all of our team members and enterprise licenses. Of course, we do a lot of work with these leading AI labs. So we have many more, but that's the one that are broadly available. And in there we teach them how to develop their own custom GPTs and how to use a dummy data and pilot some of their ideas. And then if they feel like they have onto something, they come to us, we experiment with that, uh, we start to scale it a bit more. And then of course, if it is a successful pilot, we look to our more advanced custom engineering team to help us scale that into a successful initiative. Final slide here is an example of that. We'll pop over here is we did this with our people operations team in . We, uh, we got about , people operations tickets annually. Uh, we went through a period of massive hyper growth within our organization. Uh, that's contributed a lot to those tickets. We created a custom GPT over a weekend that read inbound emails or tickets. It wrote a response based on its knowledge base. And then the human specialist, which is always our human in the loop, would review that draft and then send, uh, an answer within two days of implementing that custom GPT per ticket processing time went down . Huge productivity gain, very tangible when it comes to in the form of labor hours. And then we set out, uh, to uh, scale that of course, and I have some recent updates here. This is from this year most recently is now ai. Our AI assistant conversational language model is handling about of those tickets. We saw an interesting trend as our ticket volume actually increased. And what we found when we dug into those tickets, 'cause we were expecting somewhat of the opposite, is people were finding that the conversational language model that they were interacting with was being so helpful to them. They would tend to ask it even more questions. So we had to expand that knowledge base and refine any gaps that we had. We've also achieved record EMPS scores this year, uh, since our founding. They've never been hired and many of the comments were that, especially from an HR perspective, they were getting instantaneous answers. And then of course, they always have the ability to answer or request for a human. So Zach, uh, you know, I'll pause here. We'll pop over to questions 'cause I do wanna make this a dialogue with the time that we have left. Yeah, that'd be great. I gotta, could you actually bring up that last slide again quick? Yeah, I'd love to ask a follow up in this. And, um, this came, this question that I'm coming to you with came from a conversation I had with another member in our leadership network around how they are actively hiring and really looking to build more like AI fluency skills that they're looking for in talent. So one, they're either building that on their current talent that they have, but also all new hires. That's how they're kind of evaluating, like do they have a base level AI fluency it like in order to work here. Like that's what we look for in new hires. Um, so now that we have like a people op people operations team that maybe had a skillset before that they needed to be extremely knowledgeable and X topics, right? They needed to be able to respond and communicate effectively in response to the tickets coming in. Like, these are their skill sets. That's not necessarily the key skillset anymore if AI is answering these tickets. So have you have, we, like, have you reformatted maybe the skills that you're focused on with the people operations team in this scenario now that they're more of working with an AI tool versus actually answering the tickets. Does that make any sense? I know that was a wordy question. It's actually a fantastic sentiment and question is with any executive on this call, our l and d leaders that are on this call, this is your bread and butter. Let's just take away the term AI for a moment. This is simply a new tool, disruptive watershed moment that's entered the marketplace. It is going to change jobs, roles are going to need to evolve, and it's part of us to help identify what does that context mean in your organization. To get very specific in our organization, because our conversational language model we refer to as Alan has started to do of the tickets, that's a significant amount of labor hours that were freed up on our people operations team. And at that stage, this was only about two months ago, I started to have them come to me and say, I'm worried about my job. Do I even need, uh, do I have a job anymore because of how well this is doing? And my immediate response was, you're right in the sense that your people operations specialist role is coming to an end in this environment. But how it's gonna evolve is you're now going to be an AI engineer with a domain expertise in people operations. So that expertise act that you mentioned is still critical for the human component. Now, I needed to upskill the team on how do they start to qa, how do they start to refine? That's why I threw in here token utilization refinement for those not familiar with tokens, simple, uh, put here is compute or electricity to power. All of these tickets cost money and we need to figure out ways to refine the responses of Alan so it uses less electricity. And that's what that team is now being upskilled to learn how to do. And so the roles have evolved and we still need them on our team. We have capped hiring, but we still need, uh, those team members on our team for that particular team. Yeah. Uh, Felipe asked a great follow up question and was wondering if you had any other examples of use cases like this one that came out of the more like the employee engagement kind of driven approach where you're asking employees, Hey, what are things you would like to kind of offset or start to get AI assistance with or automation with? What are some of the other things that you've maybe seen so far? Thanks, Felipe for that question. It's a, a critical aspect of this journey, as with any technolo technology, is to have a feedback loop, a mechanism in place where you're getting this feedback from your team members. And that's not unique to ai. One of the use cases that we have is an l and D one, which I realize I should have used that example with this group. But if you go back and think of this slide here, every one of my team members on my people team is evaluating how do they end up making in the aggregate an instructional designer, for example. And we got a lot of comments. We have a global footprint where new hires that were being onboarded kept coming to us and saying, Hey, your facilitators are nice, but the accent that they have or, uh, they're, they're speaking English too fast and it's becoming a barrier for our learning. And so we set out to AI our way out of this. It's the team motto at this point. And we started to work with replacing live facilitators or facilitator recordings with AI synthetic voices. We had an initial fail. Uh, and I like to be open with this as not all of this is successful at first of people saying the opposite. Once we implemented the synthetic voices of now it sounds too robotic and this is terrible. So we started to partner with some, uh, studio quality AI synthetic companies to say, Hey, we want to get your voices into our programs. And, uh, that way it's indistinguishably lifelike. But here's the key is we started to offer in our asynchronous, uh, courses that if you wanted to have a regional dialect based in let's say Indio, where we have a large footprint, you could change the voice to have a, a local dialect, uh, accent or, or speaker be the one that does your training. So all of a sudden this comment that we received from our feedback turned into a very tangible AI use case that improved learning by breaking down some barriers, uh, which I thought was just a great inclusive case. Yeah, it's amazing. Okay, I got one last follow up question for you. Sure. And you, you're maybe a special organization because obviously you're, you're helping develop this within companies. Obviously we have a lot of, uh, other organizations that are like, they don't, maybe they just got a subscription to CHE GBT for their people, right? Like that's where they are and they're trying to build that AI fluency up, right? They, one, they're try starting to map out, okay, what are the skills and where are the opportunities for human tasks versus AI tasks that, that might be like the first mapping exercise. But then after you map that, then you go, okay, well now there's critical skills that are needed to actually bring that to life. Right? And how, so maybe I'm assuming there's still pockets in your organization where that fluency was needed to develop. What were ways that you were developing those skills with your people? Or how do you start to encourage them to kind of build those AI fluency skills to leverage AI assisted tasks or automations within their role? Um, what does that look like for you? Yeah, this is a great approach and you could tell that we have a bunch of l and d leadership on the call. This is perfect example. And one thing, I'll give three points here briefly. One, break down any barrier that AI is some type of different or scary thing that we have to upskill the organization on. It's a new tool that's out there with any new technology. You have the great opportunity to go in your organization and implement an upskilling plan that's unique to your organization. Two great free courses that are out there before you go spending tons of money on courses. And, and Turing does not do training to be clear. Uh, I recommend that you go to deep learning.ai. It is a fantastic organization. They put out amazing content. And right now, a a required course from every person on my people team and largely in the organization is a gen AI for everyone. It's a fantastic course that puts it into, uh, layperson terms and easy to understand. And that will help guide your team to think how do we start to convert, uh, these, these pieces here. Uh, the final bit is an, another piece of advice is before you go spending tons of money on AI tooling, uh, I would first, if you have those chat GPT or if you have Google Workspace Geminis included in your workplace subscription, uh, tool around with a custom GPT, see if AI actually will help in the tasks that you're planning to invest capital into before you invest that capital, have that proof of concept, go to your CEO like Matt and say, here's the results we had with chat EPT helping us augment. Now we wanted to bring in this off the shelf tool that we think will pour fuel on the fire, a lot more successful budget case there, highly recommended. And then maybe a fourth bonus, if you ever have questions, do feel free to reach out to me. I'm happy to connect on these. That was awesome, Taylor. Yes, I, uh, I'll reaffirm that for all of you. I shared Taylor's LinkedIn in the chat there, always sharing amazing content and resources, but if you have follow up questions, make sure to connect with them. Taylor, this was awesome. Appreciate you taking time out of your vacation. That's crazy. Thank you for doing that. Uh, just shows how much you're, you're willing to serve this space. So thank you so much. Anytime. Thank you all. All right, everyone, how about that, right? Like, I, I really love that we're able to dig into some more of the AI use cases a little bit, but then also convert that and talk about the skills that are needed to bring that to life. How does that augment the skills profiles a little bit? If we're totally changing what tasks and responsibilities someone's actually focused on that maybe used to be their task, but now they're actually doing more quality control of something that's doing the task. Like how does that change our responsibilities? So that was awesome. Check it out. I'm definitely checking out Deep learning.ai. That sounds like an amazing resource. So check that out as well. All right, last but not least, our final session for today. I'm super excited about this one. Uh, SOTU and I met, I don't know, months ago in, um, San San Diego. Yeah, it must have been in San Diego. Yeah. San Diego, yes. Okay. Um, I was like, okay, I know I was in California, but let me bring you up here with me. Uh, we met months ago and you were in a different role at that point. I Was, yes. And you started to kind of hint how you were leading some transformations internally. You're like, Hey, I, I I think I'm gonna be moving into this other focused role. We're gonna be working on things with AI and our talent. So I'm really excited for you to kind of bring us home on this program, especially when we think about AI and hrs transformation and how also HR L and d or the people function can help spearhead this transformation internally and what that means for skills and our talent and our people. So everyone, let's give a warm welcome to TU for being here with us. Uh, I, HR transformation lead for Zoom Info. So super lucky to have her. Thank you so much all. I'll hand it over to you. Thank you so much for inviting me here. So again, my name is Sat Sen, and let me just pull up my slideshow. There you go. You should see my size. Um, so as like mentioned, uh, my current title is AI HR Transformation Lead. So I guess I'm one of the few people who hasn't lost their job because of ai. Oh my goodness, that's a terrible joke. Um, but, um, I am really excited to be leading our AI transformation. And what means is that I am really helping our employees to get really excited and, uh, skilled in, in using AI effectively and responsibly in, in their roles. Um, so a lot of enablement, a lot of comms, a lot of change management. I also get to build my own agents, which, which is, I have to say my favorite part of my job. But anyway, um, I will be speaking more, um, from the context, um, of my previous role, which is, um, that I was leading our learning and development team. Um, and that's what I've been doing for the, for the plus years of my career. So that's when my background is. Um, and Zach asked me to share how we, uh, created carrier profiles, um, in our organization at ZoomInfo. So ZoomInfo, a couple of words about what do we do. Um, so we're a global tech company with, um, around , employees, and we help our clients find other clients. So we essentially help our customers find other customers. So we're in the go-to market space, and we use AI in, in our products as well. So we have been an early adopter in, in using ai, and that's why we started using ai, um, also in this, uh, tenures, uh, task of, um, creating these carrier profiles. Um, so, um, so maybe a little bit about what do we mean with carrier profiles, because this can be called by different names. I mean, I've done this kind of work in, in all of my companies pretty much. Um, sometimes we've called them, uh, skills profiles or, uh, we even named it once, um, a carrier navigator. So essentially these profiles or paths, carrier paths, they help our employees understand what's expected in, in different types of roles, um, and so that they can, um, find opportunities to move internally, uh, within the company. So whether it's lateral or diagonal upwards, move, uh, movements. Um, so basically, uh, enforcing, reinforcing internal rotations, uh, within the company. And, uh, they can be especially helpful when you are, um, not just finding new opportunities within your, your current function where you might have a better understanding on the different roles, but they can be especially helpful when we're interested in, uh, making a cross-functional move. And, uh, you just wouldn't even know where they begin or where, where to start developing yourselves, uh, or getting ready, uh, for, for that kind of a different type of a role. Um, and then we also leverage these profiles in, um, in giving feedback and, and coaching. And, uh, and they just essentially help you to build your path or cube, carrier cube, as we say here, um, because we do believe in, uh, not just upward, uh, mo movement, but really, uh, lateral movement as as well. So, um, I'll focus a little bit, um, on how, um, how we created them and, uh, my main focus will be on especially how we use ai, but I'll also make sure to release some time for, for questions in the end. Um, so first of all, the overarching goal with these types of profiles as, um, I, I heard somebody else, uh, saying earlier in, in this, um, in these panels or in these conversations that yes, it's, it's a really ous task. So you don't do this, you don't, uh, take this on lively. So, uh, we really wanted to, um, make sure that this, this work is grounded on a couple of parameters. So first of all, the overarching goal was to drive high performance throughout the company and really looking at it from the whole employee life cycle, starting from recruit recruitment, because we are actually using these carrier profiles in our interviewing as well. We, uh, we turned them into interviewing guides, uh, for each role. So that really helped to, uh, make their hiring process much more efficient and align those interviews, especially since, um, pretty much for all of our roles. Um, which is common in, in most companies, you have multiple interviewers. So it really helped to align who's focusing on, uh, on which types of questions, uh, in their interviewing process. And then obviously help to, uh, create accordingly, uh, scorecard, um, to make those more informed, uh, hiring decisions, uh, for, for each role. And then for purpose and development, um, these really help to clear, clarify what's expected, uh, from, from each role, and then help managers to provide feedback accordingly, uh, to their, uh, direct on on what, what, uh, strengths and opportunities they may have, um, aligned to those, uh, care profiles, um, in the role that they're currently in. And then, uh, help to drive care conversations towards, um, potential next, uh, types of roles and, and help the managers to really support that development path, uh, towards, uh, getting readier for, for that kind of a role. I always like to say readier because I don't think you can ever be, or you don't have to be ready for, uh, for the next role, but I, I, that's why I always like to say readier. Um, so, um, anyway, and then for talent planning, that's, that's, um, that's another key component of, uh, of this work. So, uh, they can be also helpful, uh, in assessing promotion readiness. It's definitely not the only factor. Um, like in most places, we, we obviously also have to have the business need in place and, uh, and, uh, the role scope and, and all of that. But at least this gives, um, a little bit of a parameter, um, to, uh, to align on what, um, what, what skills do you really need to develop, um, and to be at least be ready, uh, to take on, uh, the next, uh, level of, um, of, uh, uh, of promoted, uh, role. And then as an organization, more like from the organizational development point of view, they can really help to identify what gaps, uh, there might be in, in the overall talent in, in the org. Um, but maybe with that, I'll just move on to how did we create them. So, uh, so this is where the AI part comes in, into play. So I'll show first, what did we do at first? So this was our first go around, which saved a significant amount of time compared to not using AI in the past. Um, but we, we figured out a better way, which I'll talk about right after this. So at first we started with, uh, around like gathering input from the subject matter experts and our comp team. So what, what are, what skills, what expectations are needed in a, uh, in a given role, um, at, at an each level in a certain job, family. And, uh, we talked with them to get some role specific suggestions on what development, um, um, actions can be taken if, if somebody wants to move to a next level role. So pretty much the typical stuff that we've been doing for years, uh, even without ai. Um, and then we used AI to really fine tune, uh, what the SME input was. And, uh, before we did that, um, we had trained our AI model, um, on our care profiles and our job profile framework. And then also on Radford leveling, which we use as our external leveling benchmark. And then we also trained it, uh, to use our company values because we wanted, um, the LLM to speak our language and, um, and pull some of our values, uh, language, uh, into this as well, because we have defined our values, not just words. Um, like we have five values. We haven't just, uh, said that these, these are the values, but we also have defined what does it sound like, uh, and what does it not sound like as in like actual behaviors. So that really helped us to get to a much more refined, um, end product. And then, um, once we had, um, used AI to fine tune each one of these profiles we got, we brought them back to the, uh, subject matter experts to review them. Like, does it make sense? Did we miss on something? Is there something, um, that we should fine tune or add? Um, any other examples that, um, that are missing? And then we also worked with our compensation team to make sure that the leveling was, was still correct and did any other iterations as as needed. And then we launched them. So, so this was our initial go around, but then, um, some lessons learned. So we actually then, uh, started saving significant amount of time when we, when we turned this around. So, so instead of starting from, um, gathering input from the SMEs, which you all know takes a while, like setting up meetings and, and everybody's basically whatnot. So we actually started from, um, from an AI generated draft. So we, we, we realized that, hey, we can actually get to a good , maybe even , um, since we had already spent time, uh, to, uh, train our ai, um, or the LLM that we were using for this purpose. Uh, we had created a systems prompt. Um, I was lucky to have, uh, somebody on my team who was really well-versed with creating systems froms already a couple of years ago. Now pretty much all LLMs have that, um, have that pretty readily available, or you can even build some agents, uh, to do this. But, uh, that really helped us to get to a really solid, um, , level, uh, if we wouldn't have spent the time to train the model on, on what I just mentioned earlier. So anything that's meaningful for us, our Care bench profile framework and, uh, the Redford leveling and our company values and any other guidelines, obviously we wouldn't have gotten anywhere close to that stage. Um, but that really helped us to move much faster. Um, and then once we had that AI generated draft for each one of these profiles, then we took them to the SMEs, uh, to review them and, um, and get their thoughts and almost like just fine tuned them rather than having to start from scratch. So that really saved a ton of time. And then we essentially just removed one huge phase, uh, from this process by starting from the AI generated draft, uh, rather than, uh, gathering the SME uh, into input at, at first, and then we launched them and, um, maybe a couple of, uh, lessons learned with the launch as well. So, um, it really helped us to use the Agile way, uh, here with the launching as well. So instead of waiting for all the job profiles or carrier profiles to be completed, we actually ended up launching them by job families. So whenever way we had completed, um, a certain, uh, job families say for customers access, we would launch them for that, uh, that group, uh, and then, uh, with enablement and manager and employee guides so that we were able to move faster and, uh, at least that group was, was already able to start, uh, leveraging them. Um, and then maybe this is actually my last slide before we can move on to questions and then a couple of more, um, lessons learned. So what, what these carrier profiles are definitely not, they're not something that you can just take as, as a check the box or like a sole performance metric to, to assess whether an employee is performing or, or not, but at least they give you a really solid guideline. Um, and especially with the help of ai, you can really fine tune them, uh, with, uh, with the nuances for, for different, uh, roles. And you certainly do don't want to use them just for promotional, uh, purposes. Um, and, and at least for our organization and our culture and our values, what really works well is that we're, uh, using them to reinforce, um, not just upward movement, but also lateral and diagonal es especially cross, uh, functional movement as well. And again, it's not one size fits all, really, really anything is, but at least with the, with the help of ai, we've been able to save a ton of time from what used to be a really manual work. I remember doing this in, uh, in my first roles, uh, when I started in l and d like years ago. This type of work took months, months, months. So now, now it's more like, uh, hours per profile if, if even that with the help of ai, as long as you obviously, uh, trained it properly, um, and validate the output with, with humans. So I'll stop here, um, and happy to answer any questions you might have. All right, I'm coming back up here and, uh, yeah, I was gonna ask you to go back to, not this slide. Can you go to the next slide? Yeah. Lessons learned, so, okay. Yeah, yeah. Or sorry, the next one that with the, the four boxes? Yep. This one? No, no, the one right after. Oh, The first one. Okay. This one. Yeah. Yeah. There we go. Uh, I'm curious for the fine tuning piece a little bit. Yeah. And I think this is such a critical step, right? Like obviously getting the input, understanding your roles deeply all day, that's been like such a step one that we've talked quite a bit about. Now we want to start leveraging AI and tools to accelerate rolling this out, enhance the actual output of these things. So the fine tuning is extremely important. I'm curious of how you went about the fine tuning. So you talk about like loading company values, loading mm-hmm. Leveling, you know, is this as simple for some people that maybe are trying to do this, like the quick and dirty way, like as having like a couple like Google docs with this information on it and loading it into chat GBT and then just starting to build kind of the first draft of all these things. Yep. What is kind of like the start of fine tuning look like and what were some of the lessons learned when you did fine tuning where you're like, ah, actually it's not that fine. Like it needs to be tuned more simplified, more. What were some of the lessons learned there? Yeah, there was definitely a lot of back and forth going on and iteration, but that's the great thing with, with AI that it doesn't take as long as it would take with humans. Um, and that's why that was one of the reasons why we ended up changing up the process, because when you do that iteration with SMEs, with actual humans, that back and forth takes a while, especially if you end up having to schedule a meeting and then, uh, finding availability and, and whatnot. So, so that fine tuning that iteration with, with AI took significantly less, less amount of time. So happy to talk a little bit about how we went about it. So we use Tropic, um, um, as, as our tool. And, and back then it has, um, I mean, I think it still has, um, assistance prompt where you're able to essentially train it on. Um, uh, these are our company values, and I want you to align, um, all of our leveling based on Ra Radford, um, and we uploaded all these documents as, as Google Docs are, uh, Google sheets are whatnot, uh, into the, into the LMM. So essentially what that systems prompt does is that instead of you having to tell it as in prompt it every single time when you're creating a new profile, use these, they're already built in there so you don't have to keep on doing it. So all you have to do is, um, at the, at the role that you are creating the, uh, carrier profile for, and then it already knows the rest because it has been drained on it. Um, the only word of caution is, is just like with any AI tools, you definitely want to be mindful of what you put in. Um, is, is that that it stays in a secure environment. So we only use our enterprise licenses. So, uh, you said chat GPT? Sure, you can use that or any other LLM, but I would be just mindful that you're using, um, an enterprise grade license so that the information stays, uh, secure. Yeah, I've already heard some horror stories about data leakages and different things that cause people to be in compliance or, or even training AI on content and information that they weren't legally able to train the AI on. So there's definitely like some things to look out there. Okay. Let me move now to the launch. And I'm curious how employee employees or leaders reacted to this, right? Like, obviously you're already going through a little bit of a co-creation process by bringing subject matter experts into the creation of this, so you're already kind of getting buy-in and engagement through that, but then you have the rest of the organization that's like, okay, what does this mean? What does this mean for me? My role, new things that I have to engage with? Uh, yeah. Tell me about kind of like the launch piece a little bit more and how, um, one, what was kind of the response? I'm just curious of like, were people excited, confused, uh, a little mix of both, um, and, and how did they maybe start to move forward from there after the launch, like living and breathing, incorporating this into their work? Yeah, sure. So I think what helped us is that, uh, we already have these kinds of care profiles in place. So this was not something net new that we introduced to the organization, but what I was talking about here, we essentially refreshed all of our care profiles. Some of them hadn't been updated in couple of years. Um, so obviously for a technical roles especially, it just wouldn't be meaningful to look at something, uh, for a software engineer that was created two years ago, uh, especially in, in, in the current, uh, pace, uh, how everything develops. So, so I think that really helped with the launch, that it wasn't something that just came out of the blue, um, that they, they had never heard about before. And also since this is such a common tool to have in place in, in most companies, even with the newer employees who might have not have engaged with our previous versions of our care profiles, most of them had seen some kinds of versions in, in their, um, throughout their career in, in other companies. Um, and I think what helped is that we really, uh, customize the enablement and the communications, um, for managers and then for employees, because the stuff that we created for managers was obviously pretty detailed and, and really focused on how can you as a manager, uh, leverage these profiles? And it had a lot more information then what would have been helpful for our employees. Uh, I think if we would have tried to just do the same for all, it would've been very overwhelming for our employees, or if we would have met them somewhere in the middle, managers might have felt that I don't have enough, like I need more. Um, and, um, and then we also hosted some live sessions that were really targeted for managers and, and then, um, other sessions targeted for, for employees. So I think that helped quite a bit. And then we just talked about them, uh, in, um, in anything, uh, that was relevant, uh, or, or at least loosely related to, um, to these carrier profiles. So, um, having them visible in, in many places. And, um, um, and like you, like you pointed out, um, some of the subject matter experts who we have worked, worked on this, they also became almost like our champions in, in helping to roll this out and, and making sure that they're they're being used. Yeah. I think I, oh, keep going. Sorry. Yeah, Sorry. I would just also add, um, I think one of the most favorite parts of of this was, um, how they really helped, uh, to shape the interviews. Like I was talking about how we use them as, as the basis for interviewing guidelines as well. I think that helped to bring, um, a lot more structure than what we had ever had, uh, before in our interviewing guidelines as well. So, so that was probably the favorite part for, for employees and managers, whoever is, uh, part of, um, interviewing candidates. Yeah. Yeah. And I'll second what Shelly shared in the Chad and Champions for the rollout and throughout the entire process is so important. And I think that's for anyone doing anything, people first oriented, including the rollout and the creation of things like this, like becoming more skills based and creating these profiles like integrate co-creation into the process early on so that you can build momentum and champions and people who can kind of help spearhead the way forward as you launch and integrate this into the new way of operating. Mm-hmm. Um, okay. I'll ask one last question and then we'll, we'll wrap this up. But, uh, you talked about like having different types of guides for employees and managers and managers being more detailed. Mm-hmm. Uh, don't need the full guide insights, but I'm just curious what were maybe some of the top critical things you included in the manager guide that for some of the people who are, are listening where they're like, okay, yeah, we're gonna roll it out. Let's develop this guide or playbook for leaders to utilize this within their role and their team. What are some of like the top things that like even after feedback, you're like, whoa, we should really, you need to have this in the guide. Like these were some of the critical things that make or break it for success for that team? Yeah, so I would say keep it as practical as possible. Uh, meaning give really practical examples, like this is how you can use these in one-on-one conversations. These are the snippets that you can use when you're having a career conversation with your, uh, with your employee. Um, if you employees not performing aligned to expectations, use this. So like, literally taking it to that level rather than just giving them a huge guide on everything under the sky, because then they won't create it. So, um, that tends to resonate, uh, pretty much with everything we do here, keep it, uh, concise enough, um, and really focus on those practical examples. Like, what can you do with this, uh, when you are facing this issue? So almost like using these kinds of scenario based, um, or creating a guide from, from those types of scenarios. Like, if you are dealing with this, do this. Um, and then obviously when we were heading into the performance conversations time, uh, that's when we also, um, uh, pulled in some of that information. How can you use, uh, these, uh, these carrier profiles when you are discussing what's expected for you, um, in, in this role? And, uh, tying it into those, uh, performance, uh, conversations as as well. Um, and then also offering couple of different modalities. So we often like to create these guides, uh, which are often just really short couple of pages, Google Docs, um, and then also, um, especially when we're rolling out something new like with this one, um, also providing some live sessions that folks can tune in and, uh, then we'll always record them and then they can watch them, um, at daily show. So, so that tends to help as well that there are a couple of different types of, um, um, material, uh, uh, versions that they can, they can, um, they can read on or interact with. Yeah, I mean, that's pretty cool. It sounds like a legitimate playbook. Like if you think about, uh, you know, for me, I'm a a huge sports fan. It's football season, obviously, and I think about like, yeah, teams have plays dedicated based on scenarios that are happening in the game, right? Mm-hmm. Like if the defense is playing in a certain format, the offense has a certain playbook mm-hmm. For those specific scenarios. So how can we give our managers and our people the plays to leverage these frameworks depending on the scenarios and situations they're dealing with. Mm-hmm. And to even take it a step further, it'd be really cool where you could then take, let's say you have the company AI agent assistant, you could train it on that playbook, and that's a more interactive way where a manager could be like, Hey, I'm having this tough situation. What's a, what's an approach from the playbook that I could leverage to navigate this today? Right? Yeah. That's one of the agents that I'm actually developing right now, so, Oh my gosh. All right. Yeah. Well, I guess, okay, give me one closing thoughts as I, as we wrap up today's program. What's next for you then? Are you, it sounds like the agents and kind of more interactive assistance, what's like the big next thing that you're working on? Oh, gosh. Uh, so many things, but, um, I think the, in addition to building these agents, what I'm doing right now for certain situations, what I'm ex especially excited about, um, sooner than later, once those agents can really start interacting with each other, I think that will unlock so much. And then, um, I'm also working on together with our engineering team, uh, to build some integrations with our, um, HRIS systems and, and whatnot. So then that will unlock so much more because the agents that, um, that, um, I'm able to build right now, they're a little bit limited in, uh, in what they're, how well the numbers they are. So, so that's what I'm mo most excited about. Once, um, I will almost like unlock the next level, uh, with the integrations and, uh, and agents, uh, talking to each other, but again, we still need humans for sure. So, cool. Well, thank you so much. This was incredible. Thank you for joining this day and, uh, being part of the, our community and just sharing some wisdom and guidance from your experience. And it's awesome to hear what you're working on. So I appreciate you being here with us. Yeah, thank you so much for reminding me. And, um, anybody feel free, free to connect with me on, on LinkedIn, always willing to geek out on l and d and Onis. Yes. Thank you. Uh, as shared. Yeah, give it up first, the two for, for closing this up. I put her LinkedIn in the chat there, so connect and follow her there. And yeah, let's give her one last round applause for that. So thank you so much. Thank you.

‍