#HRTechChat: Aman Kaur-Shaik on Why AI Adoption in HR Starts with Knowledge, Data, and Culture

Speaker A: Hi everyone, this is Dylan Taggart here with another HR tech chat as part of three, uh, Hundred and sixty Insights podcast series. I'm joined today by Aman Kharshaik. She's a member of 360 Insights Global Executive Advisory Council, the current HR director at Nutrien, and, uh, a recipient of the 2024 Global Top 100 HR Executives. Um, despite her role at Nutrien, today we're just going to be speaking broadly about the trends that we're seeing. And um, Mon, thank you again for joining me. I know it's not the first time we've been podcasting, So today I know we wanted to talk a little bit about the successful, uh, adoption of AI. It's kind of top of mind for a lot of people today. I think as AI becomes a bit more ubiquitous in the market, people are realizing that education is becoming an essential part of it. Not only to get people up to speed, but also to kind of quell fear on AI. Uh, you know, so people don't think it's just coming for their jobs all the time. Um, so let's start there. So what are you seeing, you know, from the AI standpoint as a HR leader and what do you think leads to a successful, uh, adoption of AI business?

Speaker B: So this is really common, right? So the, the common conversation that we are seeing in the industry today around AI, especially in hr, is the how can we go forward? Because one of the things that HR folks, um, in the industry are mindful that the sensitive and confidential information we have and what are the guardrails around using AI. So the struggle I'm seeing that's um, around AI is not about adopting AI, but it's around, about the environment where AI is rolling out. The environment is still, I will say before 2023, but AI from 2023 to today is very sophisticated. So I think the first thing that we really looking at, um, in the industry, the conversation is about the environment we are in and the AI adoption that we really need to do. And we're struggling to say, um, the path forward or the starting point as well. So when I go to conferences and part of panel conversations, one of the questions is commonly asked is, um, how do we start? Where do we start? Uh, and because everyone is mindful of the confidential and sensitive information we have in our industry, so many teams, I think the expectations around, uh, AI also was really, really hype last year. I think this year we will see the hype is settled. Um, we are not talking about or expecting that AI will change immediate or

Speaker C: have an immediate transformational uh, impact. It's a, it's a smaller impact that the AI is making as we are

Speaker B: adopting and learning more about AI and the, the power of AI.

Speaker C: Um, so those are the, those are the key things that the conversations are happening. But at the same time let's go back to the environment that I mentioned. The environment we are in today, um, in the industry it's very common and it's not related to one organization. It's related to probably all of us can relate to that. We have fragmented knowledge, we have outdated content, we have technology that's probably designed for 1990s but we haven't catch up with the cloud mindsets. And here we are talking about gen AI and the data quality is probably poor as well. So those are the environments that AI is rolled out and we are asking ourselves to adopt which is not um, idle environment for AI to be impactful and provide us um, the transformational or the impact or the value we are looking for. So that's the struggle I'm seeing where teams or where hr, the conversation people are struggling to adopt AI.

Speaker A: Yeah, when you say the hype is settling, what uh, you know I agree with that sentiment as well. What are you kind of, what are you seeing that's making that be the case?

Speaker B: Before it was like oh we got the um, say chatgpt, it's gonna answer all the questions right? And now we are seeing is like it's answering the questions based on our policies, um, or based on the knowledge base but our policies or the knowledge basis, um, that uh, in various organization wherever you are, it's a common problem, um, is fragmented or out of date or vague. So even as HR professional you probably calling a friend to answer a question

Speaker C: or clarify that policy, then how can AI uh, uh, can answer that? So the I think first it was like oh God, this is going to solve our bold problems. And now we are understanding that AI is in at very early stages. It's infinite. Like even the AI has um, uh, the way the tools work today and the framework behind them today is very different from 2023. So in 2023 it was like you probably heard about prompting and all that. And today nobody's talking about prompting because the tools got sophisticated. So the hype around it is settling down as the reality or experimentation that we are doing with it is becoming a reality. What the tool can do and cannot do for us today, yes it will improve as we are seeing the improvements but at the same time um, the conversation is also around what we really need to do at our end for the AI to be effectively help us and augment our teams.

Speaker A: Yeah. And for a team that's struggling with adoption and not really happy with how AI is performing so far, what advice would you give to them?

Speaker B: I would just say that AI doesn't create knowledge. It amplifies the data that you have, have or the knowledge that you have. So you have to go back to your basics, uh, to get your knowledge base in order, clean up your knowledge base, make sure your policies are not vague, uh, your guidelines are there to support so that the AI can provide consistent, uh, information out or produce the information that you are looking. So AI doesn't work for us. I would say, like don't make AI do your work. Make AI augment your thinking and your strategies going forward. Um, and that's the key. But it can only do if it has access to the knowledge. So um, if, if your knowledge base is fragmented or scattered or um, is not even cleaned up, it has several versions of the truth on it, then how would AI know which, which knowledge base to. Or which article to pick up to answer the questions? Um, also the data, like what knowledge people are consuming, how they are consuming so that you can actually improve from there. You have to go back into that data as well to understand, um, who is consuming your knowledge and how, uh, the, um, how they are consuming, what are they using for, are they getting the value out of it. You have to understand that data as well, not just keep producing, using AI or uh, think that I will answer all the questions. So I will say that go back to basics. And basics are two things. Your knowledge management and your data analytics.

Speaker A: Yeah. And how do you, you know, for leaders that are struggling with that concept, you know, how do you. It's, it's a big, it's a big ask to suddenly restructure the way that data is captured. And if you don't have the research to hire someone to do that, what is, what is a good first step to kind of rebuilding or starting from, you know, the ground floor of. Because m. It's tricky for some people. You know, they, they are just using their own intuition to determine whether or not data is good or not. And which makes obviously, obviously the system inherently flawed as a human, I guess, in a sense, unless it's connected to the outside world. But you know, how do you, the data custodianship that you're kind of talking about for someone that doesn't have someone in that specific role, what way would you say for Them what would you think is the best way for someone like that to frame data so they can approach it and kind of take that data scientist approach without being one?

Speaker B: That's a great question because I had to ask that myself a couple of years ago is like when uh, I was saying oh, AI is not working. But then I realized that what do I have to do? So I call it 1% continuous improvement mindset. Right? So don't the way I approach this not solve the world problem or all the problems I have in one go, but improving 1% every day. So embedding uh, in your day to day, what you really need to look at is um, first of all anything that rework, you have to remove that rework. Ah, and you have to look at what's causing friction um, in your processes, in your way of working. Then that, that is also data, um, and also that um, how, how can AI augment the work you are doing? Right? So if you're touching the keyboard, why are you touching the keyboard? And can AI do that? If you're answering the same question again and again and again, can AI help you with that, um, answering those questions. So there are various data points that you can get. So I'm sure people will probably have um, say a uh, system, such a system or uh, such as successfactors or workday, uh, or they have service now or they are using um, teams or Slack. There's various data points where you can collect the data that how is the knowledge is consumed and what questions how the customers are interacting with you. So that's where you can start looking at the data like just data data points and start looking at what, where is the 1% improvement you can make. And then once you figure it out, the logic how to make that improvement, then you can start thinking about how you can use AI to augment that improvement and embed that AI into you, um, your day to day operations or day to day work even if you're not working into your day to day life. Right. So, so that's how I went there. Um, how would you adopt the data mindset but augment yourself your personal productivity, then productivity of your team, then productivity

Speaker C: of your organization, um, and success. I think that we have to keep that in mind. And I talked about uh, the hype and reality around AI is successful adoption of AI is not a big launch, it's a small change that you can make and repeatable habits, that's where you can start embedding and those repeatable habits or rework or repeatable Work or administrative manual work. Those are also data points within your processes that you can utilize to make some decisions around um, use cases of AI?

Speaker A: Yeah. So if you you know, agree, I feel like a question a lot of people are probably wondering about, you know, given that the trend these days, you know, post pandemic. And also as budgets for things like HR shrink a little bit, people um, are being asked to do yes, more work with less resources and AI obviously. And a lot of leaders who maybe don't quite understand how HR works are saying oh you can just use AI, figure it out for you. So, so let's um, say put yourself in the shoes or a one of a one or two person HR team at a company where, where is the first place people should start looking to implement it? Is it like what repeatable tasks? Is it in payroll? Is it in workforce management? What would you, what would you say is the best place to start looking for improve improvements and streamlining?

Speaker C: Yeah, that's a great question as well.

Speaker B: Like how would you do it? And how can you see it as um, not a technology conversation but a productivity improvement enhancement conversation. And AI is also shouldn't be the conversation about um, uh, cost cutting. Otherwise again that will become your barrier as well. So they, I would just say that um, the framework I use is the four part um, model that first of all you have to have strong knowledge foundations. So in knowledge if you're creating a policy and to roll out the policies, you have knowledge articles for example. So once you have written your policy, uh, and what we tend to do as SMB, we try to explain in the same jargon, technical way the policy for people to read. Right. But use AI to create your knowledge articles like FAQs or step by step instructions. Because AI is, I feel like AI use more human language than humans. It actually produces the knowledge articles like our FAQs in a very digestible way and in a really chunky way and it links it back to where the policy or the clause is coming. Um, so use that. That's the first place that I would say use it. Like if you're rolling out your policies or programs or guidelines, use AI to use the, to create the knowledge articles or FAQs around it. You will be surprised how many questions AI can think as compared to how many questions you can think. Right. So um, I was playing up with a policy leave policy other day and um, and when I used the AI to create faq, it was going on and on and on. I was like how many questions can you think I can think only four or five. So I think that's a great way to start because one of the things I noticed um, in my um, career is people just don't write FAQs. They love to write the policies and just think that people will just um, understand those policies because they just don't have the capacity.

Speaker C: But if you use AI, that creates the capacity in you to create those FAQs and you can have successful adoption of those policies, a successful rollout of your programs as well. So I would just say to create a knowledge and have a good solid knowledge foundation. That's the one thing that you can start with and then start embedding like AI habits on a daily basis in your um, day to day. Because first of all it has to be what is in it for me? That's the question I'm asking myself. If I adopt AI, then what is in it for me? So for me it's my productivity enhancement. So if I am creating say uh, for example, um, I have a PowerPoint and then I want to write the speaker notes for it rather than thinking how one use AI to write the speaker notes. And of course it's say it's not going to be 100% but it's 50% correct and you can refine it. You can also prepare for meetings like if you're presenting to executives, just ask AI that here's my deck, I'm presenting and um, to the C suite leadership, uh, give me the questions they will ask. So prepare for those questions upfront because AI can anticipate what questions those are. So that's also not only your productivity but it's a strategic alignment that you can use AI for and that, that will help you to change your habits.

Speaker B: It will help you to your brain to start anticipating questions as well. Start anticipating what? How the receiver of your message will see, receive that message and that you get that real time feedback um, before you actually go and roll out or present to the executives for approvals, for example. Then the data, like use the data, um, and don't get hyped about it that nobody's consuming this thing and it's no good, it's not producing. It takes time, it's a transformation, it's a marathon, it's not a sprint. Um, it goes to. I always say like for example, safety in all our lives. Safety belt is, we talk about it, we have fines every single thing. But what changed the habit of people and behavior of people wearing seat belt is that noise that cars make, like they keep going on and on and on till you put that seat belt. No fines, no taking our points of license. Nothing can stop uh, us from changing our behavior. The only that because then sound annoyed us so much and how long we are driving cars. So that's, that's why we are humans and not robots. So AI adoption takes a while. Just don't think that your people or your teams are not adopting and you give up. It's, it's a long journey and marathon. You just have to um, showcase what's in it for them and also not only with their personal productivity, your team's productivity, and then you link it with your, your company's mission or vision. Um, and then you build the capability, right? So you have to have a program

Speaker C: in place to build the capability around

Speaker B: AI, Starting with do's and don'ts. Like that is the first education you

Speaker C: have to give, uh, your people to. What don't you do with AI?

Speaker B: Uh, because you cannot blindly trust AI, you cannot blindly feed the data, the sensitive data to AI. So these are the four part program

Speaker C: that our model I have used personally in my uh, two years of my journey with AI to adopt AI successfully. So I think this four part model will work in any organization or in any teams if we go really, really basic and keep it very, very simple.

Speaker D: Hey, it's Libby Delucian. If you're sick of business advice that sounds cute but doesn't actually move the needle, this one's for you. The Libby Delusion podcast is where we go one layer deep. No gurus, no glitter, just straight up unfiltered. Talk about scaling a business, running a household and holding onto your sanity in the middle of the chaos. If you're the bottleneck in your business or in your own life, I'll show you how to fix it. Simple, tactical, and one layer deep every time. Let's go visit Libbyd.com and check out our podcast today.

Speaker A: Yeah, I think that's a lot of great advice. I'm curious, you know, obviously, 360, we're based in the United States, so we talk a lot about the United States and we often talk to people in Europe in uh, you know, you're in Sydney, in a place like Australia, or at least in Sydney, how are things kind of looking and feeling in terms of public and private sentiment towards AI and is there any sort of government regulations beginning to take effect?

Speaker B: We are very proud AI Australian government is probably the only government or the first government to have AI safety standard in place. And uh, I, I normally don't read uh, government documents, but I actually enjoyed reading that document. It's a, it's a volunteer framework that

Speaker C: the companies can uh, adopt. Um, and I can share the link uh with you so that you can share uh, when you publish this podcast,

Speaker B: Dylan, it talks about 10 tenets in there, like 10 uh, principles that we can be mindful of like AI.

Speaker C: We have to use AI responsible like the roles and responsibilities when using AI

Speaker B: or rolling out your AI. So it's a really great um, framework that uh, is available from m Australian government and I was really, really happy about that and government has focused on it. We understand that AI is the future

Speaker C: and can enhance the productivity of the

Speaker B: whole country, um, and the economy benefits to it. But at the same time we also understand the impact on the uh, on the workforce that will happen, on the economy. It will happen. So it's a very common conversation uh, that we have, we have, we used to have a minister of AI and I think we still have it. Um, so we really um.

Speaker C: In Australia we are very much talking about. So is New Zealand. New Zealand. Uh, I attended the um, Mastering SAP conference last year and they were talking about how much AI can contribute to the economy in New Zealand too.

Speaker B: So on this side of the planet

Speaker C: AI is a very common conversation, uh, both in government and in private organizations. We are, our big banks are going through huge transformation of uh, using AI, especially the um, self service, the transformation around self service so that there is a reduction, uh, escalation to say tier one, tier two to their uh, SMEs and creating the expertise in their SMEs as well. So it's pretty much a live conversation on a daily basis in our lives on this side of the world.

Speaker A: Yeah, that's good to hear. I feel like in all the conversations I hear here in the United States, it's more about you know, people feeling that there needs to be a government framework over top. But as usually is the case here, it's you know, the market quote unquote has to decided. So you know, businesses have to self regulate just simply to stay either in compliance with existing laws or to not scare off talent. And it becomes more of a reputational regulation or uh, reputations that lead to regulation. Because if you work at a company that feel, you know, is maybe doing morally onerous things with AI or support or is leaking data or is not very secure, um, people don't want to work for them or they get a bad reputation or they get boycotted and then you know, maybe in 2032 we'll see AI really regulation in the same sense if ever in the United States. Um, so it's interesting how things work. I do. You know, the more people talk about uh, the future of AI, the more you hear about something like you're talking about in Australia like those 10 tenants or like a AI constitution or at least an AI that follows the laws of whatever domestic country you're in. There's positives and negatives to that, but I do feel like it, it helps bring uh, a mutual understanding between the indoctrinated and the, and the, and those who are fresh to it because at least it, it's governed by the same laws that we.

Speaker C: Yeah.

Speaker B: And physical world. I think it's also important that government have some kind of volunteer framework for now as we are growing with AI so that we are not killing innovation. Uh, we are encouraging innovation and experimentation but with safeguards. Right. So like to smaller companies or like mom and pop shops, they don't have money to have big fours or consultants

Speaker C: uh, come in and implement AI. That's where they can rely on the frameworks that the governments can put out there, uh, and have, have that adopted if they have smaller workforce, like less than 5. But they, they want to use AI to enhance their productivity and enhance their business. Right. So um, and there, there is a, uh, of course data available like how automation helped us in, in the past, how automation helped manufacturing, how AI can help the creativity of the humans as well. Right. You, you just have to think the idea and AI can enhance your idea as well. So there's the innovation piece. But not everybody can spend a lot of consulting dollars to adopt AI or have a safeguard AI.

Speaker B: So that's where these frameworks and government has a very.

Speaker C: I, um, think they should, they should play important role and active role in safeguarding their citizens and at the same time encouraging experimentation and innovation.

Speaker A: Yeah, I couldn't agree more. I think that's a great place to leave it. Aman, thank you so much. Um, is there anything else you'd like

Speaker C: to add before we wrap up that

Speaker B: uh, keep it in mind that AI doesn't generate knowledge. So you have to have the knowledge in place and also just have realistic expectations from AI. Um, if you want to roll it adoption, it's a marathon, it's not a sprint. It takes a while to change behavior and adopt. Um, and AI is, I would say AI is not a technology challenge. It's more the data and the knowledge

Speaker C: and also behavior challenge. So organizations that invest in all three, like the knowledge, the behavior and the

Speaker B: data will be very, very successful. So as an individual, we also have to think that way, like, where is

Speaker C: my knowledge, where is my data so

Speaker B: that I can invest my time and how can I enhance my productivity, my team's productivity, my organization's or my, my community's productivity as well? What value AI is going to add to augment my thinking, my productivity? So I would just close with that thought.

Speaker C: Like, think about it in adoption of

Speaker B: AI as not technology challenge, but as more cultural challenges.

Speaker A: Yeah, I think that's great advice. Oman, thank you so much. And it's great to speak, to speak with you again.

Speaker C: Thanks, Dylan. Anytime.

Speaker B: AI is my favorite topic.

Speaker A: I'm sure it'll be relevant for a while. All right, take care, everyone. Thank you, everyone, for tuning in. And we'll see you on the next one.