Emerging HR Use Cases For Machine Learning And Generative AI

Data teams need partners, and HR has 2 use case categories that Generative AI and machine learning support better than any other technology. HR teams are challenged by cutbacks and growing demand to acquire high-end, hard-to-source talent.

In my experience, HR leaders are some of the most receptive to adopting analytics and machine learning solutions. Organizational leaders with the most significant challenges have the greatest incentives to transform.

Data needs business or customer context to be useful for analytics and machine learning. HR use cases introduce a third context: talent. Business context is the domain knowledge required to operate the business. HR is tasked with providing the business with the expertise or domain knowledge it needs to create and deliver value to customers. There’s an implied partnership with technology.

With data, analytics, and machine learning approaches, the business has an opportunity to augment its people and their domain knowledge. The current and next generation of AI-supported systems will partner with the firm’s talent to operate the business. It’s a fundamental shift that will see an increasing share of the operating model transferred from people to technology.

HR owns a key part of business transformation because Generative AI is part of the business’s talent strategy.

High-Value HR Use Cases Aren’t Obvious

This week, I was on a panel discussing Generative AI’s impacts on HR. Two use case categories emerged during the discussion. Generative AI will take on parts of HR workflows, but what that looks like isn’t apparent.

Generative Interfaces will provide employees with access to HR services. Information requests can be completed without a person being involved, but we think about these use cases too narrowly. Generative Interfaces support complete information request workflows. An employee can explore a career transition through a single Generative Interface.

“What AI product management roles are available?”

“Does my current skills profile make me a top candidate for one of those roles?”

“What gaps do I need to fill to become a top candidate?”

“What training resources does the company offer to help me fill in those gaps?”

This workflow isn’t feasible with current user interfaces. The parts that are available would require the employee to bounce between multiple applications. They would probably need support from someone in HR at each step. For companies like SAP, the core thesis is, “The more the platform touches, the more valuable a Generative Interface is.”

Microsoft is focused on the same value proposition with Copilot. It’s currently integrated into Bing, Windows, and Office 365, with more integrations coming soon.

Meta AI brings orchestration to multiple apps across its social media ecosystem. Meta AI’s Generative AI characters give the interface a personality. Meta has plans to bring that new dimension to businesses soon.

Generative Interfaces create a single access point for super apps that support complete workflows.

Separate solutions can automate the individual workflow parts: looking up jobs, matching qualifications, assessing gaps, and finding training resources. To get the full value from Generative Interfaces, a single interface needs access to data from all of them. Smaller vendors see every solution in the chain implementing its own Generative Interface, but that paradigm can’t provide the orchestration required to support longer workflows.

Resources are more accessible when employees can express their needs with a natural language prompt. In the current user interface paradigm, employees must know what’s available and where to find it. Adding Generative Interfaces to each part of the chain doesn’t improve that by much.

Not everything will fall into the fully self-service category. Recruiting and decision-heavy workflows will require a person to remain at the center, but the level of effort will decrease dramatically.

In the career transition example, the workflow takes less than 15 minutes with a Generative Interface. However, we wouldn’t take the employee out of the workflow. A career transition should be their choice, and career planning should follow their needs. It’s the same for recruiting and other decision-heavy HR workflows. The person should initiate the process and needs autonomy over each decision.

HR Planning Delivers Greater Value

The second category is even less obvious. Generative AI will change the business’s talent needs. One of the fastest-growing job requirements is experience working with Generative AI tools. Buying and developing tools won’t return value to the business unless employees have the capabilities to leverage them as more than novelties.

Using Generative AI for business is very different than playing with Bard or ChatGPT for personal applications. It’s a new workflow requiring people who understand how to prompt models to support them in delivering work products faster. This has a two-sided impact on talent planning. The talent needs are a big part, but Generative AI will also impact future staffing levels. The business needs new skills but fewer new hires than in the past.

That puts HR at the center of planning, and machine learning can support the organization. SAP is one of the first to explicitly call these new use cases out, but more will quickly follow. Talent is becoming more complex.

External sourcing alone won’t give the business access to the talent it needs. Every business is competing for the same very small candidate pool, so a new talent pipeline must be developed. Internal training and upskilling must meet current and future business needs. Talent must be developed to support future opportunities, or the business won’t be agile enough to respond in time.

Access to talent is a forward-looking challenge that requires data from across the business to plan for. Today, most HR teams only have enough data to respond to current talent needs. Developing forward-looking training and talent strategies requires access to more data.

Getting HR Off The Ground amp; Showing Machine Learning’s Potential

The data team’s first instinct is to identify use cases with HR leaders and start building. As soon as someone looks at the data, the challenge will be apparent. Most HR systems are built for historical reporting, but not for analytics or machine learning use cases, and definitely not for Generative AI. This problem isn’t confined to the HR team. Most systems were built for digital data use cases like BI, not analytics or machine learning.

Before anything can be built, businesses need to get their data in order. Significant gaps in data and infrastructure block most business units from moving forward with analytics or machine learning use cases. Why start this with HR?

Talent and training touch every part of the business. HR has relationships built with every organization, so collaboration is easier. The needs and value alignment are more evident than for other use cases.

HR owns long workflows that benefit from enterprise-wide data (recruit to retire and talent strategy). It’s an opportunity to showcase what analytics and machine learning can deliver in combination with Generative AI.

Understanding employee context and future talent needs requires data from multiple organizations. Data must be collected across longer time spans and bigger workflows. HR needs visibility into parts of the talent equation it’s never had before. That requires a new focus on data generation vs. data gathering.

There are high-value questions that kick off data initiatives.

• What makes a great employee?
• How does the business retain and lose employees?
• What capabilities will the business need next year and in five years?

HR must be aligned with the business today and in the coming years. HR strategy can bring the business’s talent needs into focus. Generative AI has a role to play, but it’s not what most people envision today. There’s an opportunity for data teams to partner with HR and deliver value.