Nichol Bradford, SHRM’s executive-in-residence for AI+HI, sits down with Microsoft’s Christopher J. Fernandez to discuss the company’s innovative approach to AI in HR. Learn about Microsoft’s intriguing 18-month journey as “customer zero” and the unique three-step process they followed that could improve the odds of success for your organization’s AI reinvention.
Nichol Bradford, SHRM’s executive-in-residence for AI+HI, sits down with Microsoft’s Christopher J. Fernandez to discuss the company’s innovative approach to AI in HR. Learn about Microsoft’s intriguing 18-month journey as “customer zero” and the unique three-step process they followed that could improve the odds of success for your organization’s AI reinvention.
Rate/review AI+HI on Apple Podcasts and Spotify.
Nichol Bradford:
Welcome to the AI+HI Project, a SHRM podcast. I'm your host, Nichol Bradford, SHRM's Executive-in-Residence. Thanks for joining us. Each week, we sit down with experts to provide strategic insights, practical tips, and actionable strategies to innovate a future that blends human ingenuity and advancing technology. This week we're discussing a strategic human-centered approach to AI adoption in HR, rooted in responsible AI principles. Joining us today we have Christopher Fernandez, corporate vice president of human resources at Microsoft. Welcome to the AI+HI Project, Christopher.
Christopher Fernandez:
Thank you, Nicole. Happy to be here.
Nichol Bradford:
I've been so excited to talk to you because of just the scale and the scope of the AI and HR transformation at Microsoft, and so you've led this transformation over the last 18 months. What inspired you to take on this challenge, and what perspectives on AI and HR has evolved during this journey?
Christopher Fernandez:
Well, myself and many colleagues collaborating together have been on this journey, and when the opportunity first arose, I reflected on my past 30 years within the HR profession. I've always had this deep desire to affect the wellbeing of individuals in helping them achieve their own aspirations, coupled with enabling organizations to create value. However, that organization defines value. Those two focal areas, coupled with what technology that's available now can enable, was very exciting to me, and it is very exciting to me. So that was my main motivation: to be able to take the professional experience I've had over these past three decades and use it in a way that enables this really fantastic technology to affect people's lives in a very positive way, both at work and, in general, in their life.
Nichol Bradford:
And one of the things that has really stood out in the things that you've written online is this human-centered approach to AI adoption in HR. Can you explain for our audience what that concept means to you and why do you think it's essential for successful AI integration and HR practices?
Christopher Fernandez:
The ability for people to use technology in service to their goals and their aspirations as well as the organization that they're a part of, needs to be central in how we think about the practical application of technology. Consider the past when we would've had an HR professional, or a finance or legal professional, for example, or anyone, for that matter, engaging with technology. The technology itself allowed us to get access to the information that's been digitized, but it didn't necessarily allow us to be able to extract it in an integrated way that allowed for insight.
Human-centered approach in this context really enables a person, the human being, to be able to distill insight and information that allows them to create more capacity in their own creativity, to be able to also remove aspects of the work that aren't necessarily of the highest value to them as people. The technology also enables individuals to gather and gain insight in a very democratized way that they might not have had access to elsewhere or, in the past, would've been easily able to get that information. So these things coupled together enable a very human-centered approach that is the driver behind all that my work's been focused in on inside of Microsoft to enable people to get that information.
Nichol Bradford:
What I love about the human-centered approach is that I think often when people are thinking about AI and they're thinking about technology, they really focus on the technology side, but it's the human side that actually makes the implementations work. And so, your integration clearly demonstrates HR's role as a strategic driver of innovation in the company, and so for HR leaders in our audience who are really being asked to become more strategic in their organizations, how has this project reinforced, or how do you see AI integrations as reinforcing HR and its role as a key business partner?
Christopher Fernandez:
If you think about technology, historically, you would've had to go through multiple iterations of learning the technology. What is the productivity tools? How do you utilize them? How to think about the operating system? How to think about the browser? How to distill all this together in a way that extracts knowledge and insight in service to your work? All of that's collapsed down now when you think of AI, with two major drivers having a natural interface. If you think back historically when people would interact with technology, think of the graphical interface or think of touch screens, for example, the advent of that. Having natural language to interface with technology requires people to think in a way that drives what their insights are with collaboration of the technology and with the technology. Couple that with just the general distillation of knowledge and insight that would've otherwise been lost in the past because they would not have been able to go through all the various sources of knowledge and insight being brought together through these large language models, these very sophisticated reasoning models.
Take those two pieces: the ability to have natural interface with technology coupled with these large, sophisticated models, all tied together to allow the person to utilize technology in ways they wouldn't have been able to in the past, to accomplish their goals and the organization's goals. HR, in my view, is central toward that end, having a complete thought about how that motion should work, having a plan of record toward that end, understanding human behavioral expectations of how individuals and ones will work with one another, and incorporating the technology. HR, the CHRO, their organization, coupled with their colleagues across their senior leadership team, drive that effort. The absence of which it is people are uncertain; they look back, and they say, "What will this mean to me as an individual human being? How is this going to impact my job and my work?" Having a behavioral science overlay to all that I just described can enable better adoption and utilization in ways that can benefit the person and the organization. That's, in my view, where HR plays a very unique role.
Nichol Bradford:
And especially in these AI implementations where so much behavior has to change, where people are thinking differently about their work, how they do their work, what their workflows look like, and to me, HR is almost like the circulatory system of an organization. Can you walk us through the process of how you identified and prioritized your HR functions that you would update? Walk us through the process of how you laid out your implementation and how you decided what was first.
Christopher Fernandez:
First, it was getting people comfortable with technology, generally speaking. Not so much in the sense of engaging with AI as we define it today, but just general automation and having people, individuals, have agency over what that technology is. So early on, we used what was commonly referred to as low-code or no-code application development, where the HR professional, leveraging their domain expertise, could engage with this low-code application development technology, take what was largely arduous and not particularly interesting work that was highly manual and automating that work. Extracting data and insights through a fairly simple interface, applying it into the application that they were building, which was very intuitive, and you need not be a technologist to do this, and then having the outcome of an application that automated their work, that they had direct agency over and control. Step two then was providing a roadmap, what I'm commonly referring to here as a business architecture, where they began to see their individual applications pinned into that architecture and saw how that improved experiences for themselves and their colleagues.
Draw that parallel over to that third step, which is knowing what to ask, having the ability to prop the right questions. I use this phrase quite often, sort of the Socratic method, the Socratic doctor, where you know what to ask because you have depth of understanding in your profession that will drive the Copilot to create outcomes that are directly in service to your work. Just like you had the experience when you're building the apps, technology was directly in service to your work where you had agency and control over it. The business architecture over this period of time, as we went through step one, step two, step three, business architecture became more and more refined, not in a vacuum, not from top down, working with communities across the HR profession together, defining where they thought the best effort was to enable that architecture that was initially laid out where we wanted to spend more effort and energy and where we want to spend less effort and energy based upon their incumbency knowledge.
Their knowledge of the work, where they are the experts. That business architecture began to grow, evolve, such that they began to see, "Wow, the work we're doing is directly impacting our ability to enable employees to have better experiences broadly in the company." And we, as individual HR professionals, are able to have a better experience in our professional aspirations and be able to focus in on the higher-end aspects of our role profiles, what we would have not been able to do in the past.
Nichol Bradford:
I love it, and I have so many deeper questions about that too. One of the ones that really rises to the top for me is: How does the way that you've integrated agency reduce fear of change? Because it sounds like your people could really see how it was supporting them in doing their work. Could you speak to that? Because there's a lot of people out there who are really afraid of the change right now.
Christopher Fernandez:
Take two examples. The steam engine, which we often talk about as it relates to an example of a technological advancement in the printing press. Imagine, back when the printing press was created, individuals saying, "This distillation of knowledge. I don't know if this is really necessary, at scale." Fast-forward to today and think of the corpus of knowledge all that we see around us in the form of books. Imagine the rejection of that concept now. It's part of our daily existence; the printing press has enabled knowledge sharing in ways that have advanced the wellbeing of all of humanity.
Steam engines have advanced the wellbeing of humanity in various ways of how people experience life and created all new sorts of interesting jobs that were non-existent prior to that point. When you think about those examples, and you fast-forward to now, and you draw on the history to say, "The technology is there to augment and to create capacity for people," think of what we can do in service to one another by having removal of the drudgery of some aspects of the work through automation, through distillation of knowledge and information through Copilot in a natural language manner to engage with technology and then create a better experience for our colleagues who are doing other work and within our organization beyond HR, by our efforts.
That, to me, created a sensibility about you do have agency over this technology, because it is, in essence, a tool. One of the great misnomers that I have seen over time, and I've expressed this in conversations with colleagues, is we say artificial intelligence. I think it's not accurate to say that, truly. True intelligence in the human sense is what you and I do each and every day, and so does the rest of humanity. This is an amazing set of tools that can affect how we go about our daily lives, both in work and beyond work, but like the printing press and the steam engine. It's not a replacement for humans; it's just another amazing set of offerings that we can harness and service to our humanity.
Nichol Bradford:
One of the things that I'd love to know more about are the citizen developers that you have, because you mentioned a lot of times when people are thinking about getting the AI fluency in the organization, raising the waterline. It's right now; it's sort of around prompting and people doing prompting. But what I think I heard you say is actually you started your fluency journey with giving people no-code tools so they could go ahead and start to automate their workflow, which I think also implicitly says, "Hey, it's okay. We want you to do this; we want you to do the higher-order things that you actually came into HR to do." So could you speak a little bit about the citizen developers and how you approach that for your phase one?
Christopher Fernandez:
I appreciate the way you framed exactly the thinking and the approach I've been trying to undertake here. The citizen developers are central to all of this. The individuals that early on developed these no-code applications effectively, they created a sense of community and collegiality in a way that only they authentically could do. Where they could share ideas with one another to this point about how do you really ensure agency over the technology and a sense of trust? That's going to occur not by me saying it; it's going to occur by colleagues and peers talking to one another and saying, "This is actually working for me. I created this low-code application and look how it automated these workflows, thereby freeing me up to do other higher-end aspects of my role profile." That, over time, has been a key learning for all of us. The citizen developer, early on, was allowed to learn, experiment, fell, and then achieve. Without, in my own experience, without having to have complete control.
I think it's so important to this notion that individual people, as is the case with most people, want to have a sense that they have an ability to can see and control their own future and their own destiny in a meaningful way. There has to be at some point to my comment about we were talking about these three steps: step two, having a business architecture that's communicated, so people then can affiliate with that business architecture, and they can coalesce around main deliverables for that business architecture, certain workflows and processes and et cetera.
But in the early stages, you have to give people room to think, learn, and grow. And, to be more precise, create an environment where they can experiment with the technology. That then is the opportunity to move into more advanced aspects of the technology, where they can see that as a... To use the term we use, a Copilot to what they do as the pilot, as they themselves are the pilot to use the expression, to follow through. The citizen developers are not only the individuals who are doing that early work with technology, not because they're technologists, but because they're domain experts. They know their domain the best; they know how best to apply workflow and work processes in service to their efforts.
Nichol Bradford:
One of the things that I love about it is it communicates such respect, to empower someone to be able to be a citizen developer. So then, when I get to see my piece get tied into the business architecture, and I get to see that the way that I've changed my work from a paddle boat to a speed boat, because I've been able to use these tools and I get to see it tied in. Then I imagine that it makes it so everyone can see when they're rowing together, but it's connected to this technology.
Christopher Fernandez:
It's sort of like the very beginning of our conversation when you said, "Why, what got you interested in coming into this space?" It really is rooted back to that notion of how do you enable the individual to achieve what they aspire to, coupled with how do you enable organizations to create value. And when those two things come together, therein lies the magic.
Nichol Bradford:
Well, so when we have the magic, what does the future of HR look like? Could you paint the picture for the HR people who are listening and wondering what their roles might be over the next five to 10 years when they have integrated AI into their organizations?
Christopher Fernandez:
I do not think and see AI technology replacing HR professionals. I see HR professionals who utilize AI having a much greater impact and ability to have influence and results, create it when they actually use AI. Over time, the behavioral science behind how humans interact with technology is going to become, in my view, and nobody. I try to be careful not to make predictions because nobody knows what the future is, but my sense is when I look down the road, having a deep understanding of human behavior is going to be central to the practicable application of AI at scale, not just at work but in life in general. Having a deep appreciation for how humans think, how they interact with one another, how they experience life. So, I see the HR role being central to helping to assess, quantify the human behavioral experience in ways that encourage engagement with technology in a multitude of different scenarios inside of work and outside of work, in some ways becomes more seamless over time. To create the right infrastructures, to create the right ecosystem.
Historically, HR has played a role where we said, "We want to recruit people. Come into an organization, we want to help them grow and develop through training and learning development. We want to incent behaviors through compensation. We want to make sure that leaders are developing appropriately. We want the right organizational structure overlay to enable people to work across different indices to create results and outcomes." All that's still going to be true, but the nature of the human experience is going to be so different with the technology. With AI, having HR have a sensibility that they can help guide leadership teams in how to position the human being with the technology is going to be paramount.
Nichol Bradford:
And so the future that you've described really sounds like HR people, as the architects of employee experience, are really unlocking the human potential of the teams inside their organization.
Christopher Fernandez:
I think HR is the future of AI. I really believe it's central, and your encapsulation of what you and I just discussed is exactly right. Time will tell if that's true. Obviously, nobody can predict the future, but we certainly can have a plan toward the future. We've all heard this phrase: if you want to predict the future, make it happen. Paraphrasing: Make it happen. So this is where our role comes into play as HR leaders in organizations.
Nichol Bradford:
That's actually the sign-off on my email is that quote, so I really get it. Well, that leads us to how do we measure success? How do we know that it's working? How do we know that people are using the technology in the best way? Could you say a little bit about how Microsoft has measured success?
Christopher Fernandez:
Yeah, sure. There's both qualitative and quantitative ways to consider measurement. The quantitative ways are what you typically would think. Say, for example, in our HR service centers, we're looking at throughput, efficiency, cycle time, response to inbound inquiries, efficiency gains that can be applied in other aspects of our work in material ways. That's sort of the, what I would say, the more empirical aspects. The qualitative aspects are also through environmental scans. Our employees giving an expression of appreciation. Do they see the value? Are they writing down in verbatims, in their feedback, that we see this changing? We look at net satisfaction scores, for example, with our HR service centers, and they've gone up. In a statistically significant way, they've gone up, and so that's one indicator that we're on the right path, but that's the exchange in cycle time, et cetera. The actual sort of human experience, what is that like?
Therein lies the quantitative complement to the qualitative feedback through the written word, and so we look at those environmental scans. We actually use Copilot to distill hundreds of thousands of verbatims to understand what people are saying and thinking about it. And we look then to see longitudinal trend lines both in the qualitative and quantitative aspects to say, "Are things improving?" And at this point, we are on a great path, but we don't take it for granted. We have to have a regular beat to understand, to get that insight, both the quantitative and the qualitative, and then feed that back into our use case development and elaborating on the use cases as they're being deployed and new ones as they're being considered, and then also make sure it pins back into that business architecture that you're trying to enable.
Nichol Bradford:
What advice would you give HR leaders and also business leaders who are just starting their journey? Specifically or particularly at organizations that might be hesitant about adopting AI? What's the advice first for the HR leader and then second for the business leader?
Christopher Fernandez:
I would say, in both cases, take a step back. Ask yourself a fundamental question: "What are you trying to accomplish?" I have used this expression before. "What are the complete thoughts that you have in service to your organization?" Having an empathy for what your colleagues are going to experience in their day-to-day work, and how can this AI technology create a better experience for them? In my own experience, having a deep sense of empathy toward others helps drive decision-making. At the outset, you might say, "Well, is it really a financial decision coupled with a return on that investment from that financial decision?" Yeah, of course there's going to be that overlay. But rooting yourself back to what are you trying to accomplish? How are you ensuring that you're helping the wellbeing of others as simultaneously achieving your organization's goals? The business leader and the HR leader, to me, are interchangeable in the context of what we're trying to accomplish here.
The HR leader needs to be considering things like, "What's the return on the investment, for investment in this type of technology? How will the employee experience change? How will the end-user experience change? Are they creating value for the organization? Are they creating value for all the key stakeholders around the organization?" In the consideration of customers, "Are customers getting a better experience?" I do not think that's the purview of the "business leader" versus the HR leader. I think they're synonymous and aligned toward this end. All organizations, as you and I, I think, were alluding to a little earlier, are a compilation of people trying to accomplish a set of goals. I see the HR leader, I see the CHRO, for example, and the CEO being intricately tied together, just to use an illustrative expression.
Nichol Bradford:
Yes, and I think for HR leaders who are being called on to be more strategic, for them to really understand what you just described is, that's the path and that's the way. Well, I so appreciate your time. It was fantastic. A big thank you to you, Christopher Fernandez, for sharing your experience and deep insights. Before we say goodbye, everyone, I encourage you to follow the AI+HI Project wherever you enjoy your podcast, and audience reviews have a real impact on a podcast's visibility. So if you enjoyed today's episode, please take a moment to leave a review and help others discover the show. Thank you for joining the podcast and the conversation, and we'll catch you next time on the AI+HI Project.