Discussions about AI in HR are widespread, but many organizations still face a disconnect between lofty expectations and tangible results. In fact, our recent survey revealed that while an encouraging 42 percent of respondents using AI have seen an increase in productivity, 27 percent have yet to experience AI’s impact.
While opinions vary on how best to harness AI in the HR space, there is almost unanimous agreement across the industry that artificial intelligence offers immense potential and equally significant challenges. In this context, we wanted to hear how several industry leaders are navigating this leap at their organizations.
Neha Sharma, Head of HR at Unifi, Ramesh Razdan, CIO at Bain, and Ralph Wiechers, EVP of Digital HR & People Operations at DHL, recently came together for a dynamic panel discussion moderated by Avature CEO Dimitri Boylan. They shared real-world insights on how companies can use AI to address the challenge of creating a competitive advantage.
Everyone needs a game plan, and we hope these takeaways help to inform yours. Let’s jump in.
The Pursuit of Opportunities for AI in HR
According to our survey, 95 percent of respondents intend to implement AI and machine learning in their HR processes in the coming year. Conscious of some of the challenges around this, Boylan was keen to talk about how our panel intended to do so. After an animated exchange of opinions, four key takeaways emerged.
1. Define a Prioritization Framework
In the face of so much opportunity, deciding where to invest and deploy AI can feel overwhelming. Razdan shared a failsafe prioritization framework that he recommends for determining the feasibility and value of AI initiatives by balancing data complexity with potential impact.
We have a framework. We build this 2×2 matrix that considers what value is at stake, whether the value at stake would be operational efficiency or whether you can build a new product or create new value for the company. But also, how is the ease of implementation, which is the availability of the data, how cross-functional it is.”
Ramesh Razdan
CIO at Bain
Progress is near-impossible until you master complex data management. But those able to deal with the data effectively will soon grasp whether a project is worth pursuing. Once you’ve taken the time to plot your potential use cases on the matrix, it becomes easier to focus on those high-value, easy-to-execute projects.
Boylan was quick to agree with Razdan: “Frameworks is an area that HR has to focus on to make sure they don’t end up just with a lot of AI scattered throughout a tech stack… Put that in a framework so that you have a benchmarking capability, you have guardrails so that people know what they need to know about what other people are doing, and so that the whole thing hangs together at some point in the future.”
2. Focus On Low-Hanging Fruit
Sixty-one percent of the audience shared that they have already deployed an HR-specific artificial intelligence use case at scale. The result is promising, though there’s a whole lot more to do.
From Sharma’s perspective, there is plenty of low-hanging fruit that will help the HR organization go from good to great with AI. For example, she believes that AI will help her solve real HR pain points, like improving how to conduct employee engagement surveys. Instead of forcing employees into boxes through rigid rating scales that do not reflect how people truly think or feel, she envisages letting them express themselves in their own words, before using Natural Language Processing to analyze sentiment at scale.
What are the low-hanging fruits that can truly give you a lot of impact with less effort? That’s in the short term. I think that’s where all of us are trying to leverage AI to really change the way we operate and create value for different stakeholders.”
Neha Sharma
Head of HR at Unifi
Retention is another major focus area at Unifi, and an AI-driven attrition-prediction model using multiple data points is already providing valuable insights for Sharma’s team. She shared that this model currently runs at 50 percent accuracy. Though she would like that number to be much higher, it’s good enough to start using and tweaking and for incremental improvements to be integrated over time. In this regard, she advocates for progress over perfection.
3. Balance a Top-Down Mandate with Grassroots Innovation
According to our panelists, AI-driven HR transformation should involve both grassroots efforts and strategic top-down guidance. Each shared that their organization is embarking on different levels of controlled experiments within different parts of the business to truly assess the real value that AI can bring them.
Razdan shared that Bain’s innovative culture lends itself to this approach, and bottom-up innovation in the form of hackathons has uncovered unexpected, high-value AI innovations while generating employee engagement. He acknowledged that failure is part of the process—successful organizations test quickly, iterate and move forward. At the same time, when it comes to setting an overarching strategy and working on “those big hairy problems,” top-down guidance is best. DHL shares a similar perspective.
The approach that we have taken is balanced between grassroots on the one side and a kind of top-down guidance, but not too strict, on the other hand. What do I mean by this? We quite quickly came up with our own set of compliant ways to use large language models… But, on the other hand, to not limit use cases too much. We have grassroots activity to tap into the exploration and the knowledge of our advanced teams that want to work with it.”
Ralph Wiechers
EVP of Digital HR & People Operations at DHL
4. Build Flexibility into Your Strategy
We are at the dawn of the AI revolution, with advances coming at us on an almost weekly cadence. In this context, Razdan stressed the importance of investing in a flexible AI architecture that will allow for changes in direction. Boylan agreed, sharing that this was the motivation for designing Avature’s Models-as-a-Service architecture for AI.
We are in the early innings, and technology and architecture are evolving so rapidly. So, building a flexible architecture that you are able to adjust as the needs arise is also important.”
Ramesh Razdan
CIO at Bain
AI Literacy: The Right Amount for the Right People
The conversation swiftly moved on to the theme of AI literacy. When it comes to educating the workforce on AI, our panelists were clear that a one-size-fits-all approach won’t suffice.
For both DHL and Unifi, whose workforces are made up of between 80 to 90 percent frontline staff, there is a clear distinction between those that really need to know about AI and those that don’t.
Blue-collar workers are increasingly exposed to AI in their workflow, such as the routing tools that support DHL drivers in their delivery routes. However, Wiechers highlighted that often, these employees might not even know that they are interacting with AI. In this regard, he shared a pertinent observation. Ensuring that AI is intuitive for this group of the workforce is far more important than educating them on AI. Such is their design that many of the AI tools now emerging are actually more intuitive than those that they are replacing.
A good example is picking robots that help people in a warehouse. In the past, you might have had onboarding training for those who wanted to co-work with the robots. The new technologies that are building on AI and Gen AI are so super intuitive that there’s even less need to train because it’s very personalized and very intuitive to work with, which is a bit counterintuitive. Gen AI or AI can even reduce the training needs.”
Ralph Wiechers
EVP of Digital HR & People Operations at DHL
Sharma echoed this perspective, highlighting that she is more focused on change management. If the solutions she and her team develop are simple enough, frontline employees will seamlessly adopt them as part of their daily lives. That being said, there are other user groups that require a much higher level of literacy, given the nature of their work – HR being one of them.
Boylan cautioned that some basic training might be a good idea for everyone to reduce operational risk, for example, ensuring employees aren’t putting intellectual property into ChatGPT. Furthermore, recent headlines provide a stark reminder that AI doesn’t value the truth, and it might be good to educate the workforce on that.
I stress that it’s not an oracle of the truth. It’s a statistical engine that gives you the most popular combination of words in a sequence, and it can use the truth, but it’s not obligated to produce the truth. So, there are some fundamental things that people need to understand, based on exactly how they’re interacting with it.”
Dimitri Boylan
CEO at Avature
At the technical helm of a very different kind of organization, Razdan shared that Bain offers training to 100 percent of its employees and now includes Gen AI training as part of their onboarding programs. As a management consultancy that supports clients with their own responsible and ethical artificial intelligence strategies, this level of education for its workforce is something that differentiates Bain from a value proposition perspective.
With every organization falling somewhere on the scale between Bain and DHL or Unifi, defining how much time and effort to invest in increasing AI literacy across the workforce will depend on the nature of your business. However, increasing AI literacy across the HR function is non-negotiable.
There is a dire need for all of us to really understand the pros and cons, challenges and risks that AI presents and see how best to leverage it. So, there is a level of understanding that is needed as the function, and that holds true for some of the other corporate functions as well.”
Neha Sharma
Head of HR at Unifi
The Future of AI in HR
AI represents a sea change in technology. It’s a prediction engine, a consumer of a huge amount of context. No system has ever been able to incorporate this degree of context when producing an answer. It’s also the most powerful knowledge transfer tool that mankind has ever created. With these fundamental components to draw on, how does the talent function use this technology to elevate its value inside the organization? We sought to answer this question with our panelists.
As technology itself is evolving, the role of talent function is to enable everybody to operate at their full potential. That’s to me, what the talent function of the future is in partnership with technology.”
Ramesh Razdan
CIO at Bain
Transforming Strategic Workforce Planning
A lack of robust, real-time data has handicapped HR when it comes to strategic workforce planning. As Wiechers explained, “[Workforce planning is] so granular and so dependent on local talent markets that a company can’t do it as a one-size-fits-all model. So, from my point of view, the reason why it hasn’t been cracked yet is with respect to the availability and the structure of data.”
With the right architecture in place, HR teams could leverage AI for comprehensive analysis of historical and market data, as well as predictive analytics. But alone, it’s no silver bullet. Sharma highlighted that “There is a need to upgrade the way HR systems work so that they are a lot more intuitive and cognitive in terms of the way they capture information.” Instead of requiring employees or teams to enter information manually, she envisages a near future where data can be captured on the go as workflows progress or the employee lifecycle moves forward.
When data is collected in the flow of work, some organizations will probably figure out how to conduct workforce planning in real time. This would be a game-changer for Unifi, operating as it does in such a dynamic industry.
Maybe twenty people didn’t show up and I have five more people leaving. I have X number of new hires walking in today. Then I have maybe five possible flight delays that I can anticipate today because of the weather. It’s just too much to keep up with. We are managing it today, but I think this is truly one area where AI can exponentially help.”
Neha Sharma
Head of HR at Unifi
Wiechers is serious about the need for HR to get its house in order in terms of available data. Only then can a redefinition of the business model take place.
Focus on people data as a product. If [data] is not done right and not available and not accessible, prediction models won’t work. Use cases will come by themselves; models will become commodities. That’s all easy; but if you don’t have fuel, your engine will not start.”
Ralph Wiechers
EVP of Digital HR & People Operations at DHL
Realizing the Skills Opportunity
Until now, technology has been unable to handle the complexity of people data necessary to empower a skills-based approach at scale. Addressing the recent leaps in machine learning and AI that now enable us to consider the role of skills as a currency, Wiechers and Razdan emphasized this as a key area of opportunity for HR. Of course, this is intrinsically linked to the first opportunity we identified. Enhanced visibility of skills is an essential element of successful workforce planning.
Every organization I look at has a skills gap. We have always tried to figure out how to infer the skills. We don’t really know what skills are needed. How do we leverage this technology to infer the skills and build them automatically? We human beings are generally lazy. We don’t want to do anything manually. Anything we can do automatically, systematically inferring the information is the way to go.”
Ramesh Razdan
CIO at Bain
Improving the Employee Experience
Beyond helping HR shift from service provider to driver of strategic value, our panelists also highlighted some areas in which AI could help the function better serve its employees. On the one hand, Sharma shared her vision for introducing hyper-personalization in the context of career pathing, rewards, recognition and benefits. Razdan also sees the opportunity for AI to help with information overload, a common pain point amongst enterprise employees.
There is so much information, but people can’t find information anywhere. We have information silos and how do you bring information with context? There’s a fantastic opportunity between the talent and technology functions to build that, to give the right information to the right people.”
Ramesh Razdan
CIO at Bain
People are nuanced, and that complexity has long challenged talent teams and HR technology. Boylan closed the session on an optimistic note, suggesting that AI in HR might be the gift that finally allows talent teams to really get a handle on the human.
The conversation with our panelists made one thing clear: the future of HR will be defined by how effectively organizations can align AI with their people strategy. From prioritizing the right use cases to driving AI literacy and capturing real-time workforce insights, success requires both a strong framework and a flexible mindset. At Avature, we’re helping leading enterprises put these principles into practice with an AI-powered platform designed for strategic HR transformation. If you’re ready to turn opportunity into measurable impact, get in touch with us today.