Michael Dunne and David Francis discussed the importance of talent intelligence and AI in HR and talent management. Michael highlighted the Eightfold AI 10-year journey in AI-based talent intelligence, supporting 155 countries and 30 languages. David, from Talent Tech Labs, emphasized the need for AI to inform, not decide, talent strategies. He shared data showing 52% of organizations are expanding into new industries, and 44% face top challenges like retaining talent and managing skills. AI’s role in talent intelligence was detailed, including its ability to predict, synthesize, and inform decisions. The discussion also covered the democratization of AI and its impact on workforce agility and automation.

Introduction

• Michael Dunne introduces himself as the Director of Product Marketing at Eightfold AI and the pseudo host for the meeting.
• Michael mentions the availability of the slide deck and the importance of sharing both the recording and the deck for future reference.
• The focus of the meeting is on talent intelligence and the Eightfold AI AI-based native talent intelligence platform.
• Michael highlights the importance of understanding talent needs and applying AI insights to support automation and data usage across the organization.

Talent intelligence and AI framework

• Michael discusses the concept of talent intelligence and the ability to dynamically understand talent needs across the talent lifecycle.
• He emphasizes the importance of using purpose-built AI insights and platforms that support automation and data integration.
• Michael mentions the need for a framework for Responsible AI from day one, including certifications and compliance with standards like ISO GDPR and ISO 42001.
• The discussion includes the importance of transparency, explainability, and mitigating bias in AI algorithms.

Eightfold AI background and capabilities

• Michael provides an overview of Eightfold AI, a Silicon Valley-based firm in its 10th year, specializing in AI for talent management.
• The firm supports multinational firms, public sector organizations, and philanthropies in 155 countries and 30 languages.
• Eightfold AI enables step changes in talent acquisition, management, resource management, and contingent hiring for various industries.
• Michael emphasizes the firm’s role in helping employees, job seekers, and employers make the right decisions about employment.

David Francis’ introduction and research overview

• David Francis introduces himself as the Global Head of Research at Talent Tech Labs, a research organization focused on talent technology.
• David explains the organization’s mission to help improve talent functions using technology and their unique practitioner-led insights.
• The meeting is structured to cover talent intelligence, AI trends, and the future of AI in talent management.
• David shares data points highlighting the macroeconomic drivers behind talent intelligence and the top challenges organizations face in implementing talent intelligence strategies.

Talent intelligence definition and use cases

• David defines talent intelligence as the process of collecting and analyzing people data to improve talent acquisition, development, retention, and talent strategies.
• He explains the various systems that generate data on the workforce, such as HRIS, learning tools, talent marketplaces, and compensation platforms.
• Talent intelligence platforms aggregate this data and add an AI layer to help talent practitioners make better decisions.
• David emphasizes that talent intelligence should span the entire talent function, not just talent acquisition or talent management.

Challenges and market for talent intelligence

• David discusses the challenges organizations face in implementing talent intelligence strategies, including data integration and organizational context.
• He highlights the importance of AI in talent intelligence and the need for talent intelligence platforms to be able to integrate with multiple systems.
• David explains the different personas and use cases for talent intelligence, from C-suite strategic decisions to employee development.
• He shares an example of an organization’s talent intelligence function and its mission to analyze and translate business needs and support workforce planning.

Skills and talent intelligence

• David explains the importance of skills as a foundational data source for talent intelligence platforms.
• He discusses the need for skills to be enriched with organizational context, such as industry, experience, and proficiency.
• David shares an example data flow diagram for using skills to inform job architecture and job design.
• He highlights the different types of skills initiatives, from skills architecture and hiring to skills-based compensation and workforce planning.

Impact of AI and agents on the workforce

• David introduces the concept of agentic AI and its impact on the workforce.
• He explains the evolution of AI from bespoke models to large language models and the democratization of AI access.
• David discusses the landscape of AI infrastructure companies and vendors offering agentic solutions.
• He shares data on the adoption of AI and agents in the workforce, including the use of generative AI tools and digital workers.

Final thoughts and Q&A

• David emphasizes the importance of talent leaders having an informed opinion on AI and agentic solutions.
• He discusses the potential risks and benefits of AI in talent management and the need for responsible AI practices.
• Michael and David address questions from the audience, including the impact of AI on small and mid-sized organizations and the importance of taking small steps in AI adoption.
• The meeting concludes with Michael thanking David and the audience for their participation and confirming the availability of the slide deck and recording.

Michael Dunne 00:00
I am really glad to have joined us today. David Francis, Global Head of Research at Talent Tech Labs, a long standing research house that is then specializing and expert and renowned in the area of technologies as they pertain to talent, I’ll be kind of the pseudo host today. My name is Michael Dunne. I’m a Director of Product Marketing here at Eightfold AI. We already had a question asking if the slide deck will be made available, and my understanding is yes to that question, I do strongly believe, with like 90% certainty, that we will be sharing the deck, because we very much appreciate you joining us today. There’s a lot of exciting research, there’s a lot of great ideas. We’ll be taking polls. We’ll be trying to take questions from the audience, and so we want both the recording and the deck available, because we’re really we’re here to help you in these times, these times, it can be exciting, but also maybe a little bit like, don’t know when it comes to how we can apply these innovation and technologies around the area of work, HR and talent. Now our focus here is on, excuse me, talent intelligence, and a full help pioneer the AI based native talent intelligence platform. We’re actually in our 10th year. And what I want to say here is what our perspective is on talent intelligence, and the big thing, I think, is the ability to dynamically understand your talent, your talent needs, what the market offers for needs across all stages of the talent lifecycle. And so what this comes into play means having ability to apply purpose built AI insights, have a platform that can support automation, utilizing data that could be across the organization market data, as well as ongoing data captured from your operations. And finally, data that comes with working with supporting the AI, because when you have AI, you’re also having data. The other part is just to leave for folks, along with our perspective about what talent intelligence means is also what does AI mean? And we’re seeing a lot of change. We’re seeing a lot of concerns out there. There is talk of regulation, and little more than talk of regulation, say, like that the European Union. And what we’d say is, when you’re looking at this space, and people keep saying, AI, AI, AI. And I had someone joking about that when they went to HR tech, they said, All I heard about was, AI is to see and take into account your thoughts about having a framework in place for Responsible AI from day one. So at full with our talent intelligence platform, that’s what we did. We worked with a lot of the certifications and compliance that many of you may know from the ISO GDPR for Europe. But then there are others, such as, now the ISO 42001, which is around AI management systems. How are AI products developed? How they consume resources, how it’s used? Others often work with, say, like government agencies, in terms of how your cloud business works. But along with compliance, along with audits, it really should be thinking is, if we’re bringing the AI from day one, are we really using the right data so that we get objective results? Are they the right algorithms that allow transparency, explainability? Are we using the right evaluations? Are we able to do the valuations at scale? Are we able to have other parties do it, and then, are there the right products in place for mitigating bias and ensuring objectivity? I think everyone wants to say, when we have recommendations and matches, we want to be fair and objective as much as possible. So it’s a little bit of a crusade. Of mine was that with AI and AI driving things like the talent intelligence platform, is also the thing from the start, being responsible AI by design. Now with regards to my organization, we are now in our 10th year. It is a Silicon Valley based firm that started with AI experts and teams for helping essentially both the employees, job seekers, citizens, et cetera, find the right career in the world, and on the flip side of that is helping Navy and employer networks make the right decisions about employment. So we support multinational firms, public sector. We’re active in 155 countries. Support over 30 languages. By language support, we don’t just mean the UI, we also mean parsing. We also mean understanding the skills and how those languages are used, whether it’s France, Germany, Japan, etc, and also being involved in 19 countries or industries. I mean 19 industries. But the main purpose is to enable a step change in how people approach their talent concerns, whether it’s talent acquisition, talent management, resource management, contingent hiring, etc. And it enables step changes with very complex organizations, whether you’re multinational, whether actually you’re a public sector organization or a philanthropy, and regardless of the industry that you’re in, whether it’s industrial, whether it’s in finance, whether it’s in life sciences, etc. So just kind of wanted to give a quick snapshot of a fold where we came from and where our thinking is as it applies to town, intelligence platforms, responsible AI, and how this could be applied in various markets, industry segments, and the type of outcomes that we hope to support. Following this introduction, I actually would like to hand over the soap box so to speak to our esteemed guest, David Francis, and let him then be able to share a lot of exciting research that they’ve been capturing over the past six months.

David Francis 06:25
Thank you so much, Michael. Thank you, appreciate that. Hopefully you can hear me. Okay, hello everyone. Thank you for joining us and delighted to be here and join you all today. Before we jump into the content, maybe I’ll just spend a minute to let you know who I am and who our organization is. So you know, I’ll be talking for, you know, the next half hour or so about, you know, some of the trends we’re seeing in the space and thoughts on, you know, strategy and the technology market. And it’s always useful, I think, to have a little bit of background on what’s informing that lens or that, that particular perspective. So I’m the Head of Research Talent Tech Labs. Been there about six years. We’re a research organization focused exclusively on talent technology, and so we’ve evaluated, at this point, you know, somewhere in the neighborhood of 2000 different vendors in the space. We have three practice areas, started in talent acquisition and now service talent management and the contingent workforce. But I would say kind of our unique, special sauces that were exclusively focused on, you know, work related technology and all of kind of our research and insights, we were actually founded by practitioners, and kind of our insights and our research and kind of how we how we deliver, are kind of designed or crafted through the lens of the practitioner. So, you know, our mission is to help organizations kind of improve their talent, function, be a better use of technology, and kind of everything we do, you know, is kind of around that. So hopefully that gives a little bit of perspective on us, and with that, we can kind of jump in the way we’ve organized our time here together. We’re going to start with a bit of, you know, what I’ll describe as a talent intelligence primer, where we’ll work through kind of some, some, you know, kind of high-level drivers, some background information, kind of definition of stuff. How do you actually define talent intelligence? And then, you know, where does it show up inside of an organization’s tech stack, or where might it show up inside of an organizational tech stack, and where might you go to kind of purchase these kinds of capabilities, and what are kind of the implications for the broader, you know, team and organizational infrastructure. And then we’ll look at some kind of persona-based use cases etc. We’re going to shift gears a little bit, but talk a little bit about kind of what’s coming down the horizon, or what’s around the corner with AI, and in particular, agentic AI, and what that means for kind of the organization at large, and how this is coming to market in what it means for propellant practitioners. And then we’ll end kind of time training with a bit more fireside chat discussion with Michael, and happy to answer any questions that you all might have as well. Feel free to post questions at any time. I think we also have a couple of holes, kind of baked in here, so hopefully that all that’ll go smoothly before we jump into kind of a deep discussion of talent intelligence in particular, I thought it would be useful to just share a couple of data points, you know, kind of highlighting the macro-economic, maybe drivers behind this type of solution, kind of as a category this year is from our friends over at PWC. This is a CEO survey that they ran last year. And, you know, besides being what I think is a beautiful data visualization, it also illustrates a pretty interesting dynamic, which I think is true or most organizations are experiencing right now. What they did is they asked organizations in their survey that if you know, provided the primary industry that they operate in, they asked, what new industries are you or what other industries are you expanding into? The net, net of it is that, you know, more than half of organizations are expanding into new industries outside of their primary or historical industry. And so if you’ve ever heard the sentence, you know, everybody is becoming a technology company. Well, this data gives some credence to that statement. As you can see, the number one kind of inflow in industry is technology. But this is true across the board. And so kind of across the economy, you know, car manufacturers, or traditional car manufacturers are now getting into electric vehicles. You know, technology companies themselves are oftentimes, you know, vertically integrating or expanding into the industries which historically, they’ve sold to. And so I think the kind of net here is that the pace of change is accelerating. And, you know, when organizations are expanding into new businesses, into new markets, and they need to hire and staff those businesses and markets and figure out, you know, what their workforce mix is, or maybe they’re shutting units down, all of that lends itself to being able to understand, you know, kind of workforce dynamics much more agilely. And talent intelligence is a tool to help support those types of decisions. The second data point is actually from a survey that we ran. Did this earlier this year in partnership with unleash, where we surveyed more than 200 280 organizations, talent leaders. We asked a lot of different questions. Were surveyed or were they a research company? So we love surveys, but one of the questions we asked is on the organization’s or leaders’ top challenges. And if you look at these top four here in particular, I think it’s notable that you know all four basically are challenges that you know lend themselves to being solved via talent intelligence type capabilities, and so retaining critical talent applicant quality the economy, or, you know, responding to business uncertainty, which I suspect many of my US counterparts might be dealing with right now, with the recently signed Executive Order on each one these, if my phone call and email is Any, any indication, and then keeping up with changing skills requirements, and so again, each of these are areas that are in part addressed by tone intelligence capabilities. The last data point here is from our friends at McKinsey, which did a pretty robust analysis looking at basically trying to quantify the economic impact of talent strategies, and in particular, of suboptimal talent strategies. And the main point to make here is that you know, talent strategies, or talent intelligence, isn’t necessarily just kind of like a feel-good exercise you know that you go through, but there’s a real kind of economic value or cost associated with either getting this right or getting it wrong. The net, net of the study was that they were leaving something like half a billion dollars in productivity on the table due to sub optimal talent strategies. And they quantified these in four areas, lack of skills, lack of engagement in efficiency and attrition and vacancies. Submit net here. Your talent strategy matters quite a bit. How you tackle that matters quite a bit. Getting It Right has a real economic value to it, and talent intelligence can contribute to making sure that that talent strategy is correct. All right, with the preface out of the way, now we can actually dive into our discussion of talent intelligence in particular. And what I’ve observed is that talent intelligence is one of these terms where, if you ask, you know, 20 different people, what does it mean? You tend to get 20 different answers. If you ask a vendor what it means, it tends to be, you know, what am I? What do I? You know, what product do I have to actually sell? If you ask a practitioner, you know, it can be all over the place. So this is our definition, which I’ll share here and I’m going to use to kind of frame the rest of the conversation. Talent intelligence is the process of collecting and analyzing people’s data, which includes things like skills, capabilities and even things like experiences, talent ambitions, in actual kind of work activities and work output in order to improve talent acquisition, development retention as well as inform better talent strategies. It’s a little bit of a mouthful, but I tried to kind of illustrate it here in the bottom. And what this is illustrating is, in most organizations, there are many different kind of talent systems and player people systems in play that are data generators, which means they’re generating data on, you know, on the workforce. And so the highest level you have is, you know, your HRIS, which is typically your people’s system of record. You have learning tools, which might be, you know, learning management system or learning experience platform, which capture, you know, metadata on the courses, and also data on the individuals that are taking those courses, how well they’re doing on them, and possibly even some, you know, kind of ambition type data you have, talent marketplaces or internal career sites, which might suggest directionally, you know, where people want to move, internally, possibly compensation platforms, performance measurement tools. So all of these different systems are generating data. And you know, what a talent intelligence platform does is it basically aggregates all of this up into, you know, together. It puts an AI layer on top of it to help, you know, talent practitioners make decisions, or make good decisions or better decisions using that data. And if you know, just listen to the, you know, the different systems kind of I talked about, you know, or that are represented here, what you’ll notice is that they are siloed in one particular function. And so talent intelligence shouldn’t be siloed just in talent acquisition, and it’s not just a talent acquisition capability. It shouldn’t just be siloed in talent management, and it’s not just a talent management activity, but activity, talent intelligence kind of as a function really should span, you know, an organization’s entire kind of talent function and capabilities. Now, where these different systems that generate data kind of come from, you know, obviously, are across these different silos, but the idea here is that you’re aggregating all of this information up and then making it meaningful to you know, the practitioners of the team that needs it, kind of regardless of where they sit in the in the talent function. And now we have a chance. Here we go to our first polling question. So with that, with the definition out of the way, we can use this as an opportunity to get a chance to let you all respond. Where is maybe a self rating here? How much is your organization’s use of talent? Talent intelligence technology. Give you a moment to respond to this, and then we’ll move here. And it looks like nobody is using it, or the question didn’t come through, all right, so not one person has any experience with tele intelligence technology. Well, if that is true, then this is wonderful. This should be a good educational opportunity for you. One more question, which is, what’s the biggest challenge an organization faces in implementing talent intelligence strategies? There might be an issue, because I don’t think I see anybody actually submitting. So, yeah, anyhow, I guess something will have to be figured out. If those responses come through, maybe we can go back or send them as a post follow up to the webinar here, one of the shows just to talk through. So if you think from an organization, like, like, a buyer’s perspective, you know, where do you actually like to go to market to get kind of talent intelligence capabilities here sold kind of on the concept of, like, Okay, this is interesting, but where do you actually go to get these capabilities? This is our talent acquisition technology ecosystem. It is one of three that we produce, which is kind of our lens on all the different solutions that are available to be purchased by talent teams. You know, this is the talent acquisition one in particular. We organize it by stage. So this kind of mirrors the flow that a candidate goes through in a journey from source, engage, select hire. And then if you look at the bottom, there are these kind of color codings. And the color codings are what we call verticals, which are you know sets of related solutions? Which are you know tools that kind of have a common DNA, even if they don’t necessarily, kind of compete directly with each other, and then inside of every bubble, or sub vertical, as we call it, are companies that tend to have similar feature sets, similar products and features, and tend to show up in competitive RFP as an example. If you look at the light blue area, we actually have a vertical for talent intelligence, and we include in their matching systems and labor market intelligence. And I think by extension, I would probably also say that, you know, hiring analytics and people search which are basically kind of external people search engines would be here as well. And the common DNA that these platforms share is that they typically are using AI trained on a, you know, fairly substantial amount of data in order to either service insights or inform decisions. And so with matching, it’s a decision of, you know, how well a particular candidate’s background, skill set and work history match to a particular organization, you know, job function and role that they’re hiring for. Or, you know, the labor market intelligence case, it’s, you know, kind of more macro, labor market, supply, demand, trends, compensation, skills, skills intelligence. Now it’s not as simple as this, even though this is already kind of complicated, because how the market has developed is that there are many solutions which basically have a, you know, kind of a user facing product, which is in a different category, but they’ve also added on or built in, you know, elements of talent intelligence to the platform. And so as an example, you know, a company that sells a CRM, you know, might have tele intelligence capability as part of their product, even though, kind of what they’re selling is a CRM, or maybe on the talent management side, there’s a company that’s selling an internal talent marketplace, you know, but they also, you know, as kind of part of their full platform, it includes talent intelligence capabilities. So all this is to say is that these solutions right now are kind of sold on the market, tend to be in relatively, you know, kind of a verticalized fashion, or to kind of cover a piece of talent intelligence. But it’s not necessarily like that, you know, kind of telling intelligence that we talked about, you know, talked about, you know, a couple slides ago, where it’s covering kind of the end to end, you know, talent function. It tends to be more kind of a discrete or bespoke piece of talent intelligence for a specific use case or function. So with that, maybe I’ll give you a couple of thoughts on, you know, kind of guiding principles towards, towards navigating the solution space here a little bit, and key considerations, particularly as you’re getting down into kind of evaluating specific vendors and, you know, all that fun stuff, the first, you know, maybe call out, would be the ability for A system to integrate or read data from multiple systems. As we said earlier for the more kind of data a system can take in, ingest and make sense of, the more value it’s going to be across use cases. And so you know, generally, if you were trying to have a system of talent intelligence. Generally, the more systems that you can integrate with and read data from and have that data become kind of meaningfully useful to you, the better the ability to use kind of the same system, or intelligence and engine for multiple use cases. Generally, the more that you can capture under the same umbrella, typically, the better that I’ll say, we’ll look at, kind of the persona-based use cases here in a bit. I don’t know if there’s a single system that’s going to do kind of everything for every stakeholder today, but generally, kind of, the more you know, bundled you can get one of these solutions to the extent they can actually execute against those use cases that you’re looking at, the better. And then last is the ability to kind of calibrate to organizational context. So we’re going to talk about a kind of AI, particularly here in just a moment. But for these systems to be useful oftentimes, you know, every organization is unique, and every team is unique, every individual is unique. And so there’s this kind of contact, this or this concept of, you know, needing to use AI to kind of almost automated scale, but then hyper personal, or, yeah, hyper personalized, kind of locally. And so what that means in practice is, you know, kind of any system you you know, are taking a look at, you should understand its ability to kind of calibrate the organizational context or make, you know, decisions based on specific organizational context, as opposed to something that’s just, you know, maybe coming off the off the shelf, and then the last we’ve already, kind of talked about already, but just the fact that talent, intelligence as a function spans, you know, all three areas and shouldn’t be kind of, you know, tackled in a vacuum. So this is probably a good chance to pivot a little bit to a quick talk of artificial intelligence. And I’m going to spare everyone the technical details here, and, you know, kind of the 70-year history of AI and how we got to where we are today. I’ll spend a little bit of time later talking about how yellow lenses are different from traditional models, and what that means. But for the time being, I just want to give kind of a high-level overview, kind of how AI is using these systems, why it’s used and what it should be used for, and maybe what it shouldn’t be used for. So, first and foremost, first and foremost, like, like, what is AI and what’s it? What’s it? What’s it kind of good for every you know, form of AI, from, you know, the most basic, you know, simple algorithm, to kind of the most advanced kind of leading, large language model on the cutting edge of kind of theoretical research, basically all do the same thing. They’re prediction engines, and so if they’re taking some input and then making a prediction for you know what should be, what should come next, what work should come next, what decisions should be made, etc. And so what it’s really good at is inferring relationships and then learning over time. And so as new data is fed into a system, one kind of byproduct of AI based systems is that they tend to improve over time with use, with calibration and with feedback. They’re typically praying, pre-trained on a large data set and then fine-tuned on an ongoing basis. And then, you know, I would describe them as a decision orchestrator. And what I mean by this is, if you’re thinking about a kind of talent intelligence at scale, like, actually, AI is kind of like the only way that you can really do this at scale, which a lot of organizations have done. Many organizations now have talent intelligence functions, and they have teams of, you know, data scientists, which are, you know, basically, maybe have already done step one of kind of aggregating data, and then they’re doing different types of queries, you know, against that data in order to try to answer questions that they’re getting from their business leaders, their talent partners, and that is one way, one approach you could take. But really, you know, if you want to do this kind of at scale, it’s kind of the only way to do it. And the reason is, it’s good at a few different things. It’s really good at predicting AI is really good at synthesizing information that is, you know, going into a large kind of data set and really simplifying down to kind of the cliff notes version of it, so to speak, it’s extremely good at discovering relationships or making inferences, particularly for data points that might be so disparate, you know, analyzing billions of different data points like that, you wouldn’t even be able to find that relationship kind of existed, or know that that relationship existed, and good at informing, and informing is unique from deciding which is an important distinction. And so in our view, it’s possible, in fact, that some organizations do set some kind of risk tolerances at which they’ll let you know. Ai driven session. Ai driven systems make decisions, you know, based on different, you know, kind of decision criteria, again, risk thresholds. But we would say, as a best practice, organizations should use AI to inform decisions, but not to make decisions. And so as a human is kind of involved in the loop making some type of a talent decision, an AI can inform that decision with additional data, additional context and additional insight. But since we live in a regulated industry, and I would say in an increasingly regulated industry, you know, it probably should not be deciding any information without human oversight or input into those particular decisions. The last point here is because kind of AI is kind of a, you know, capability is so important to talent intelligence capabilities, broadly speaking, if you are kind of going to market and evaluating, you know, talent intelligence platforms, or talent intelligence capabilities, and you’re looking at different vendors, you know, understanding the kind of ins and outs of how its AI works is actually critically important. It’s arguably one of the biggest, you know, maybe deciding factors, in addition to the kind of specific use cases enabled, and obviously the user experience. But this is a critically important capability for talent intelligence, and so your intelligence, and so understanding how a specific vendor does all these things under the hood is incredibly important from a decision support perspective. All right, let’s give this another shot and see if this actually works this time. How mature is your organization’s talent function regarding AI, give it a few seconds, and it looks like, again, I’m not seeing any submissions come through, so again, I’ll just assume the polling questions are working quite As expected. All right, I wanted to provide this here as a different lens on talent intelligence in particular, this is what we’ve seen, kind of in practice, how these capabilities tend to, you know, kind of either boil up in an organization or perhaps trickle down. Tends to be based on, you know, kind of the persona or the end use case that’s, you know, being designed, or the different stakeholder that’s, that’s kind of being designed for, sold to, and at the kind of, at the fundamental level, the same kind of intelligence engine could power all of these different experiences that we see here on the slide. A lot of it just tends to be on kind of, you know, product focus and, you know, vendors’ decision on what kind of market and how they want to go to market and what market they kind of want to serve. But in theory, the underlying kind of AI or talent intelligence infrastructure that powers, you know, the leaders or strategic level could also power the kind of employee experience as well. But what tends to happen is that different personas kind of want different things. From a talent intelligence platform, from a capability perspective, for the C-suite, they tend to be interested, obviously, in kind of the most strategic stuff. And so, you know, they’re interested in, how do we compare to our competitors? You know, are we in the right markets? What market should we be moving to? Where should we be expanding? Do we have the right skill sets, kind of holistically and to win today and in the future? Do we have the right talent strategy overall? Are we optimally using our talent? And so the, you know, the kind of the quote, unquote products that are, you know, designed to serve this particular persona tend to be a lot more kind of strategic in the analytics focus, so things like internal and external analytics, and labor market intelligence and kind of work for, you know, possibly like workforce planning tools. One layer down, you have kind of CHROs and talent leaders that are still interested in the strategy but a little bit more, you know, kind of, how do we operationalize this stuff and start driving actual talent strategies? And so they’re interested in things like, what are our workforce needs? How are we going to fill those gaps, you know, how do we become an employer of choice? How do we identify high potential talent, you know, which might be internal or external. How do we bring in that right talent, you know, externally? And how do we Up skill and re skill our workforce, which is becoming kind of increasingly important as job descriptions are changing as a result of AI taking over maybe some of the more automatable or mundane parts of our parts of our work. And so this tends to be, you know, an analytics flavor here too, but also kind of an execution component of it. And so it’s, you know, the quote, unquote products are things like workforce planning, skills, analytics, intelligent recommendations. And then kind of the systems of active activation, like the ATS, the CRM, vendor management system, etc. The next two are, you know, the manager and the employee. And this is a very poor kind of human, human level lead. And so managers, you know, are most interested in their scope of review of the world. And so, you know, what skills do they have on themselves or on their team? What skills do they need on their team? How are they going to identify people that they may be hiring for, either inside the organization or outside the organization? How can they help their employees grow, you know, personally, in their role currently, and also into their next role? And how can, you know, teams be organized better to actually get work done. And so from a product perspective, it’s things like succession planning, Team analytics, opportunity, marketplace, learning, assignment and hiring tools, and employees largely interested in their own kind of development in relationship to the employer. And so it’s things like, how does the employee find their next opportunity. How can they develop themselves internally and grow professionally? Who can they connect with at the organization that you know might be a mentor? How can they upskill themselves? And so these are things like learning, navigator, career exploration, you know, intelligent job search, and then mentorship and community. Now there might be some companies that are saying they have, you know, all of these things in one platform. I actually don’t think that is the case yet. I don’t think anybody has kind of fully cracked the nut across the board. But from an organizational perspective, kind of thinking through what the use case is, or the driver for, you know, some of these capabilities that you’re looking at can help kind of inform, you know, which particular kind of category of tool might be better which particular individual then vendor based on their kind of relative strengths or weaknesses in any of these particular areas. All right, it was the last time I can move my software. I forgive myself. There we go. Lest we think this is all kind of pie in the sky, you know, theoretical ivory castle theoretical exercises. There is like this is kind of a real trend, as I said, already, many organizations already have or are building out a talent intelligence function internally as part of their overall talent team. Many are investing in technology to support that function. What I want to call out here, this is an example of an organization’s talent intelligence function where, I think is nice here is they’ve kind of articulated the overarching mission of the function, and then kind of the core kind of areas that it’s going to influence, or decisions that it’s going to influence. And so the, you know, the mission of the TI function, you know, things like analyze and translate business needs, identify, identify critical skills and talent gaps. You know, support workforce planning. You know, contribute to the annual operating plan, develop function specific talent insights. And then if you look at kind of its relationship to the broad organization again, you’ll notice that it’s not just in talent acquisition, not just for hiring strategies or, you know, better sourcing tools, etc. It’s not just in talent management for, you know, slightly better, you know, internal mobility programs. It’s really across the entire organization, from, you know, the HR team to talent acquisition, even insight into mergers and acquisitions, and that’s how that’s going to, you know, influence workforce design, or design and strategy. And so, you know, maybe, I think one of the nice things that this particular organization has done is actually defined kind of a mission statement, or driving kind of purpose for the function. And, you know, it’s easy to get kind of starry eyed about shiny technology, but as you’re thinking about telling intelligence, you know, something you can do as an organization is write down kind of a mission statement for yourself and a set of kind of north stars that are going to be important for you to accomplish, which, again, having that kind of framework and roadmap in place is going to make it much easier as you’re going to market and assessing kind of different solutions that can contribute to, you Know, an ecosystem or technology provider that’s going to support that particular function. All right, I’m going to spend a couple of minutes on skills here, and this is not meant to be a skills presentation, but the reason is, skills is an incredibly important part, or like, one of the most important part of the kind of, the foundational data sources for many telling intelligence platforms is kind of a, you know, atomic unit of data that is used, kind of in many different systems. And so I want to talk a little bit about, you know, just basically how to maybe think about skills in a more effective manner than maybe they’ve traditionally been leveraged at organizations. So maybe what I want to make here is, if you think about, like, what is a skill, it’s really just a piece of text, and so it’s a string of text, like, if I, you know, was, say, searching for somebody I wanted to hire, and I put in research, as you know, a skill I was looking for, this doesn’t tell me, like a whole lot, it doesn’t have, you know, doesn’t tell me what type of research that somebody I was looking for was doing its you know, was it academic? Was it a private institution, a public institution? Was it quantitative or qualitative? Was that, you know, a market research firm? You know, there’s, there’s all these kinds of things that, you know, aren’t included along with this piece of text. And so the main point I want to make is that skills in a vacuum are kind of meaningless. And for skills to, you know, have value or be useful in different decision making. Kind of engines or paradigms, really, skills need to be kind of married to some kind of organizational context that kind of enriches, you know, what it is that you’re actually saying. And so in this particular case, you probably also want to kind of know, you know, the industry that that skill is associated with your type of organization, the experiences that you know generated that skill, and probably a proficiency anybody even want to know a little bit about, like, what are the that particular candidates, like, you know, aspirations, and in an internal context, to the extent that you had this data already, you know, maybe even some data associated with some of the outcomes or performance of the use of that particular skill. And so one of the, you know, I guess, kind of maybe things to think about as you’re evaluating talent intelligence platforms is the extent to which, you know, maybe new data or context can be added, or the extent to which a tool that you’re evaluating can actually kind of like, understand and adapt to you, You know, your own kind of unique organizational context, because just skills kind of by themselves in a vacuum, in my view, leave a little bit to be to be desired. Just another illustration of kind of how this all might come together in in this particular case, showing kind of an example data flow diagram, you know, possibly for, you know, using to actually kind of inform job architecture and show it’s showing kind of different sources of skills, where those are coming from, different systems that they might ingest with, to kind of give an overall sense of compensation and how some of these things relate to each other. You know, maybe different competencies or that you might have internally, you know, unique to your own organization, alternate career paths that might be associated with a particular job title, etc. And then one of the shares this year, kind of, as it relates to skills, there’s been, you know, for the past several years, there’s been quite a bit of, I would say, kind of organizational momentum or interest in skills, and not without merit. There’s, again, it’s kind of the foundational, you know, data point powering a lot of different new solutions. And the way it’s come to market, there’s kind of a broad variety of solutions. And so what we’ve done here is kind of organized by do they tend to be kind of organization wide, integrated solutions on the x axis, versus kind of more point solution focused, and kind of, what’s the probability of success of any of these particular initiatives? And initiatives, you know, from kind of low success to higher success. Want to be clear here, this isn’t like super duper scientific, but it’s meant to be illustrative of, like, what we’ve seen at clients that have pursued, you know, many of these strategies and initiatives. You know, in general, kind of what’s been more successful or less successful, and where we’ve seen kind of most success is in the in the upper area, things like, you know, using skills for things like skills architecture, skills-based hiring, job design skills-based learning and development and upskilling programs. And then in the lower section, you know, there’s been some, you know, I think, you know, not, there’ve been some interesting examples of companies that have done, you know, great work in these areas. But, you know, maybe a little bit more challenging to roll out kind of universally on things like skills-based compensation gig marketplaces. You know, skills-based workforce planning, which kind of necessitates an entire kind of rethinking of, you know, traditional kind of job design and you know how hiring and workforce management actually gets done. And so probably, because they’re just a little bit bigger in scope, tend to be a little bit more challenging to fully operationalize as well. Last slide I’ll leave here on skills related to, you know, I think ultimately, when we look at talent intelligence and when we look at, you know, kind of skills as part of that, ultimately, I think what organizations are trying to do is increase workforce agility. And it gets back to kind of the very first slide we had, where, you know, the workforce is changing kind of faster than ever, and the economy is changing faster than ever, and so ultimately, we’re trying to improve here as organizations. Is workforce agility. We see that as a function of, you know, understanding the jobs and work to be done, using technology to disintermediate kind of the task, and understand the skills and tasks that are, you know, required to actually do that work, and then redesigning it in a more optimal fashion via some combination of the below. You know, the kind of Lego pieces you can build, which is basically to leverage your workforce internally. Top skill or reskill you can buy, which is basically to go out to the market and either hire people or recruit them into your organization. You can borrow, which is, you know, the using contingent labor in some capacity to particularly short-term needs, or maybe longer term needs that you’re kind of building capabilities around. You can buy, which is basically automating the workforce, which is using automation, and AI, you know, to automate, historically, manual processes, or you can bundle, which is basically, you know, offshoring different parts of work to different global delivery centers, you know, around the world, and so maybe setting up teams in different geographical locations. And given number four here, but we think is going to become increasingly relevant in already having an impact on the workforce. We thought maybe it’s a good kind of transition point to talk a little bit about the impact of AI and agents on the workforce. We’ve just published a, what I think is a groundbreaking report on this particular topic, looking at AI and agents in the workforce, covering, kind of the vendors in play, the adoption, and kind of our thoughts on client strategy. We don’t have time to go through the whole thing. And so we’re going to kind of quickly run through some of the high level findings with some strategic takeaways to think about. And I don’t believe these are working, so I’m going to go ahead and skip those, but here we are. So what are our agents? This is a concept that didn’t really exist 18 months ago or 18 months ago, and now this is seemingly all we can hear about. Now, I’ll start here with maybe just a high level, kind of a gross oversimplification of, kind of how, you know, kind of today’s AI or agents are different than, you know, the AI solutions that we’ve had, you know, kind of historically and with kind of differentiating how this has evolved, kind of quite rapidly, actually, historically. All of AI, you know, broadly speaking, not just in the talent space, but broadly speaking, were bespoke models, which basically mean you would train an algorithm on some specific data set, or, you know, maybe multiple data sets for one specific purpose. And in the talent space, you know, I use text do as kind of my quintessential example. Actually I love text, by the way, it’s a great company for those of you who don’t know. They do job description optimization, particularly to make sure that your job descriptions are more inclusive and you’re not inadvertently knocking out underrepresented candidates. But you know, the way that they built that model is they trained it on, you know, lots of, lots of job descriptions, and, you know, trained it for this one particular use case, large language models, which launched around 2022 kind of changed the game, because instead of needing kind of a specific model for a specific purpose, they have kind of one model that does everything, trained on all of human knowledge, and it’s kind of a general expert across domains. And the net, net result of this is that it kind of democratized access to AI, and it made it significantly cheaper to kind of leverage AI for a variety of different applications. And so one of the reasons we’re seeing kind of an explosion of new solutions and new vendors is because now, you know, there’s much lower access to kind of the greatest AI that you know the world’s ever known at a relatively affordable price point. Now the latest evolution of this is what we’ll call a kind of multiple multi-modal AI workers or agents. And the kind of distinction here is these are, this is basically a capability that’s built on large language model kind of technology, but is unique in the fact that it’s typically given systems access, and so it kind of can operate inside of the systems that everybody uses to actually run their business, and it can actually take on work. I think of this as kind of like AI or conversationally enabled, like RPA. If you think back to like, what kind of traditional RPA was where you’re kind of automating school chair tasks, kind of agents or AI is doing a lot of that work, except enabled with this kind of conversational AI capability, which now has kind of broad domain expertise across many different applications. And, by the way, has the ability to do things like, you know, natural sounding voice and video. That’s kind of a high level of what these are. This shows kind of a breakdown or a little bit of the difference between AI and agentic AI. I’ll spend maybe just a minute here running through kind of how the landscape is developed and what it means from a practitioner’s perspective. And I think it kind of illustrates the evolving landscape of these tools fairly well, fairly succinctly. At the bottom level, you have all of the foundation model builders. These are the companies like, you know, open AI and Google and, you know, meta, that are building, you know, the world’s best large language models, investing 10s of billions of dollars in, you know, training these models. In the middle layer, you have what’s called AI infrastructure companies. And these are organizations, that basically have kind of no cold agent builders, or they leverage the foundation models, but then they make it easy for organizations to develop or deploy their own agents, whether that are vendors themselves, like, you know, eight fold, or whether that are organizations that kind of, you know, want to take a more active role in developing agents themselves. And then at the top you have all of the different vendor kinds of point solutions that have an agentic experience, which have been built, you know, using some combination of, kind of the bottom two rungs of this particular staircase. And the way that this is organized tends to be kind of by application, a point solution. And so what we’ve observed is these have kind of coalesced in a few different areas. Ai, recruiters, which tend to be sold to talent acquisition teams, which Automate, you know, parts or all of the recruiting process or talent coordination process. AI, workers, which take on, you know, segments of work, which tends to be verticalized by occupation and AI coaching, which tends to be sold to talent management teams, which is for you know, kind of role based or AI based simulation for internal, internal workers. So with that, like, where’s adoption at right now, as a survey that we ran again earlier this year, where we asked organizations about their current adoption and their expected use of different kinds of AI. We asked about, I’ll share two data points here. One question we asked about was just general, kind of like copilot and you know, like, are you giving your employees access to generative AI tools, and here, more than half of organizations had done it. Another nearly 30% were planning to roll it out in the next 12 months. And so what I’ll say is that, you know, giving employees access to a kind of a generative AI tool or a copilot at this point is kind of table stakes. And I would describe this as probably, like small T transformation, where the way this has been rolled out in many organizations is, you know, the capabilities approved, and then it’s kind of given to employees, where it’s like, Okay, you go take this now and figure out what to do with it. And there have been some, you know, some productivity gains on the back of this, and some great use cases that have developed from it. But it hasn’t necessarily, like, transformed or transformed organizations, and if you’re thinking that you’re, you know, extremely innovative, because your employees have access to, you know, an AI tool to help them, do, you know, some piece of their work, you know, unfortunately, that’s not the case anymore. We also asked about digital workers, augmenting, replacing team members. And so this was getting at, you know, to what extent are people using AI workers or AI workers or recruiters, etc, and this is, we think a much more kind of transfer will be a much more kind of transformational, you know, trend that’s going to happen over the next few years, about 24 just under a quarter, 24% reported that They had either a pilot or something active today, with another just over 40% planning a pilot in the next 12 months, we’ve had the opportunity since we’ve kind of published this, we’ve done some additional work with a number of clients. We’ve had, you know, kind of six months of labor data to observe, and what we’ve seen is there’s already kind of several like, we can already kind of see early impacts of this particular solution segment coming to market. And the big question that organizations kind of are struggling with right now is, there are so many options. There’s kind of an arms race from the vendors and towards, you know, kind of winning in this agentic space. And so from a practitioner’s perspective, there’s, there’s friction and understanding, you know, what capabilities you know, should we should we use, and which specific kind of processes should we use? Where do we get those capabilities from? And then kind of, the overall, the overall, overall arching, kind of governance structure for these solutions as well. All of this, you know, kind of has yet to be kind of formally figured out at most organizations, and so it’s kind of a work in progress. But what I will say is that we think this is coming, and I would prepare, and I would encourage you all to kind of, you know, go into this with kind of an eye towards this is happening, rather than not happening, and having an informed opinion. Because, you know, scary as this might sound or might be in some respects, you know, probably the worst outcome is not having any opinion and kind of totally outsourcing all the decision making, to say, the IT department who’s more than happy to run and make kind of all those decisions on your behalf. So there’s going to be some, you know, pretty significant implications for kind of the workforce overall, how talent is managed, for how people are re skilled, you know, for early career programs, for, you know, what specific parts of the actual like, say, recruiting process or talent management process gets automated versus which pieces stay human. And so, you know, as talent leaders, because you have kind of an informed opinion on the staff. And so I would just say to you, to, you know, get yourself educated and have an informed opinion. Because this is, this is coming, probably, whether, whether we like it or not. All right? And with that, that wraps up my discussion. So Michael, I’ll see if there maybe any, any, any any questions, or we’ll hand it back to you.

Michael Dunne 52:35
David, thank you very much for that thorough walk through today, in terms of innovation, the AI, talent intelligence, and how that’s all brought together. And I think what we’re seeing a lot of the excitement and thinking out there with global, 2000 organizations, even public sector, mid-market, etc. I do think, uh, one thing that came up was that AI, there’s resistance to it because there’s a feeling of risk aversion. Maybe it’s the biggest reason holding companies back. So do you have some thoughts on that? I know I kind of gave our preamble of how, from the start, we believe in responsible by design AI, but yeah, we did want to see maybe you have some additional thoughts on that. We’ve always talked about that. I always say, Hey, you got to get it right and then build on that. That’s part of having the platform is being the responsible AI piece.

David Francis 53:32
Yeah, don’t disagree. I think it’s a good point. What I would say is that, what I’ve observed is organizations, or many, many organizations tend to kind of like lump together in this, like, under AI, or what people think of it as AI is actually, like this really kind of broad a morph is category which includes a lot of stuff which isn’t really AI. And so if, from a risk aversion perspective, it kind of depends on the specific risk that you’re talking about. The first thing that you want to make sure of is just that like, is like, is this actually AI that we’re talking about, or is it, you know, a vendor trying to hype up what they’ve got kind of as AI? And then the second piece is for the specific use case that you’re looking to use it for. You kind of want to know, like, what are the, what are the legislation you know around that? And you know most, most larger companies that are selling to corporates that they’re to larger corporates at this at this point, have already gone through some type of an AI bought us a I bias audit test and do so kind of on a continual basis to make sure that they’re, you know, they’re there. Their algorithms are kind of inadvertently contributing to, you know, discriminatory hiring decisions as an example. But you kind of just want to look at the use case, I would say, over time, like, I feel like AI and hiring. So my hypothesis is, like, over the next five years or so, like, and I think the kind of stuff that’s happened with large language models has kind of been accelerating, like, I think that AI is just going to become kind of a part of how talent functions are run. That’s something that’s going to vary depending organization to eight organization like, in which specific areas. But I kind of liken it to driving cars. If you look at the data in aggregate for self driving cars, like, I forget how much more performant they are, but it’s like an order of magnitude safer than driving as a human. If you measure it by like, you know, whatever it is, deaths per 100,000 miles or accidents per 100,000 miles. And so at some point, it’s almost like, if you were, you know, far enough down the road, metaphorically, no pun intended, if you’re far enough, kind of in the future, you know, if, the if, if it becomes so much better at doing some of these things. Then, then a human then, you know, there’s almost a question of, like, why would you kind of put yourself at risk? Almost like, the bigger risk is, is not leaning in, um, with hiring solutions kind of in aggregate, I don’t think we’re quite near that. And so, you know, being thoughtful about how you evaluate the vendors that you’re using in the use cases that you’re that you’re doing, are very important. But if you look at kind of the best in class, AI, you know, versus what’s physically capable in a human only kind of function, you know, I think even today, the scales are already tipped towards AI being, you know better?

Michael Dunne 56:31
I agree, it is just like there’s so much data. What could be in terms of having effective where the rubber meets the road with your talent strategy, and having it work, just as we continue the road metaphor here. And I think actually it’s a great point. I think with many people, if there is resistance, I think there’s always looking in the long term, but taking small steps. I mean, because actually, as I said earlier, the AI that you’re using there should be in place around the algorithms. The data is the ability to audit it, see the explainability, see the results you’re getting. And also, I think, understanding if it’s learning like that’s part of the AI is, is the system actually learning for the basis there? I think the other question that came in, which is interesting, if I interpret this right, it sounded like, Is this all for big companies, or does business size make a difference in terms of projections, I think, like, for example, our smaller organization, our smaller size is more, should they be hesitant about using this, or should Is it, is it likely to be insured about using this in the future? So I’m interpreting that. Are we talking about this for just big companies? Is there an ecosystem out there? Is there opportunity there for what might be a small to mid size or smaller organizations? I don’t know if. Yeah. I think I’d say whether, like a 5000 FTE and under type of organization, right?

David Francis 58:01
Yeah, I think the, you know, I think the need So, when we got back to talking about, like, your talent intelligence function, and what that should look like, and kind of defining the north star in the mission, if you know, I think the capability, broadly speaking, is going to be valuable, kind of, no matter what size you are, right? Because every organization is making talent decisions all the time. It’s just that the kind of scale at which you’re making those decisions, and probably, like the specific issues, are going to be probably different if you’re, you know, a 100,000 global organization versus, you know, a 5000 person organization, you know, somewhere in the Midwest. And so I do think it’s broadly valuable. It’s just that the specific kind of things that are most valuable, or pieces of capability that are most valuable, are probably going to be different as you get bigger or smaller.

Michael Dunne 58:51
Yes. And I also think your chart earlier back I don’t know if it’s the equivalent of the ascent of man, but I think one of the elements that you’re mentioning was these capabilities are being more democratizing, like that was one reason of the pickup for the generative AI was suddenly just your employee can go to a site and start asking questions. So I think that these types of capabilities will certainly be available. I think also, people have worked around these issues. I think earlier with AI, a big thing was getting enough of the data to start having to produce the kind of results you want. I think there’s a lot of approaches now where the training sets don’t have to be a giant, big black box Lab project for a giant multinational firm. There’s a lot of work with the models and the data is underneath that could make that available to a broader audience.

David Francis 59:47
I think it’s significantly easier to deploy for a smaller organization. These days, you used to have to go train on like you’d find a population internally that you’d want to train on, and that only made sense if you had enough people internally, which kind of precluded it from being sold to watch customers. And Michael’s point that’s not necessarily the case anymore.

Michael Dunne 1:00:08
So, yeah, it’s a, I think 10 years ago, those types of projects were often described as consultant software, not the sound needed a lot of you need a lot of data scientists effort to help tweak things and interpret and then when you’re like, Wow, this is great. Let’s scale it. It’s like, it’s kind of a circumscribed lab project. Any final words of thoughts, I think we ran over a little two minutes past, but David, any final thoughts?

David Francis 1:00:37
Yeah, I’ll say thank you to everyone. We actually had a couple of, I think there were a few people that had shared questions kind of in advance. I thought they were great discussion topics. Let’s take some time to get to them, but happy to know, maybe shoot a thought or response offline. Yeah. Thank you to the team for having me, and thank everyone for joining. Hope this was valuable, and hope it’s useful as you’re making decisions around your own team’s talent function.

Michael Dunne 1:01:06
Thank you, David, and thank you everyone for joining us. We ran over a little bit, but we very much appreciate all the questions and engagement here. This session will be reposted for you to access. And as I said, a check, and I think I’m 90% sure that there will also be a deck available for you to have all the content review in terms of what David went over and the research here that’s been done at Talent Tech Labs. Again. Thank you. I appreciate it. Have a great rest of the week.