Is AI Taking the Human Out of Human Resources?

Companies recognize the benefits of machine learning for their people management teams, but do the costs outweigh the benefits?
headshot of author Mitchell Telatnik
Mitchell Telatnik
Expert Contributor
July 10, 2021
Updated: July 14, 2021
headshot of author Mitchell Telatnik
Mitchell Telatnik
Expert Contributor
July 10, 2021
Updated: July 14, 2021

Deep learning and AI are drastically changing industries such as healthcare, financial services and retail. Many companies welcome these new technologies. However, Human Resources (HR) departments are having a harder time integrating intelligent systems into their workflows.

HR departments manage organizations' employees —hiring, firing, resolving disputes, payroll, benefits and more. Many of these tasks seem ripe for automation but handing over the reins to machine learning models poses interesting ethical challenges.

AI Uses in Human Resources

  • Hiring and Onboarding
  • Employee Scheduling
  • Workforce Analytics

 

Hiring and Onboarding

The hiring process is laborious and expensive. From reviewing resumes, to interviewing and training new employees, onboarding new employees can carry astronomical costs to organizations—not to mention the new employee’s salary and benefits package. Making the wrong decision can cost even more money if the employee is a poor performer or the new hire isn’t a good fit for the position—then the process starts all over again. Not only will the organization incur the new-hire onboarding costs a second time, but there are also the peripheral costs of lost productivity to consider.

As a result, many companies have looked towards deep learning to reduce the cost of recruiting and hiring employees while increasing the quality of new hires. As we might expect, these attempts haven’t always gone as planned.

From 2014 to 2018, a team at Amazon built systems to review applicants’ resumes in an effort to streamline the process of recruiting top talent. In order to train their algorithm, the team compiled a training data set using resumes submitted to the organization over the previous ten years.

Amazon hoped this system would drastically reduce the time it took to identify the most competitive applicants by automatically identifying the top x number of potential employees. However, these engineers discovered the system was favoring male over female applicants. This was because more male than female job seekers submitted resumes to Amazon, creating a biased, skewed data set.

Creating unbiased automated hiring systems can be a difficult task. Since most companies rarely have exact gender parity, the model will identify factors it thinks are most telling of a good hire but are actually illegal for hiring managers to consider. If you’ve hired more men than women in the past, your algorithm will think men are more qualified hires than women. It all comes down to the data you’re feeding your model.

In order to create accurate hiring decisions and candidate ranking systems, we must take care when assembling the data sets to eliminate unwanted bias. For example, you may want to consider hardcoding your model to disregard certain features such as name, gender and race.

Read More From Our Data Science ExpertsSkewed Data Is the Problem With Your Statistical Model

 

Employee Scheduling

HR departments (or people teams as many companies are starting to call them) that manage hourly employees have a daunting task when creating schedules. When your employees are part- or full-time shift workers, scheduling conflicts will inevitably arise. Because of the unpredictable work schedule, a key function of HR managers (and often general managers) is managing time off and shift change requests.

If you ask any restaurant or retail store manager, scheduling and it’s related tasks often take up a large portion of their work day. However, deep learning systems are starting to take on this burden.

Automated systems can analyze employees’ requests and automatically approve or deny them based on predefined business rules. For example, many organizations that employ part-time shift workers do not allow their employees to work more than 40 hours in a given week. If a shift change request puts one employee above 40 hours for the week, the system would decline the request without any human intervention.

These systems become even more powerful when combined with predicted demand information. Accurately predicting when a particular shift requires additional employees and adjusting schedules or time-off requests accordingly can improve employee management efficiency. For instance, a restaurant’s automated staffing algorithm can learn about trends in diner traffic over time. The restaurant can then save on labor costs when demand is historically low and can also ensure adequate staffing when demand is high.

Do these technologies create a new age in productivity or simply remove the human from HR?

While these systems can drastically improve employee management and reduce the workload for managers, it could hurt employee morale. Often, time off and shift change requests can be personal in nature. If an automated system denies time off for an important event, the employee may grow resentful towards the organization and begin looking for work elsewhere. In today's tight labor market, that's a cost many employers shouldn't be willing to take.

For automated scheduling systems to work, it’s important to thoroughly define business rules and ensure there’s a way for employees to appeal the model’s decision through their manager.

More From Our Machine Learning ExpertsArtificial Intelligence vs. Machine Learning vs. Deep Learning: What’s the Difference?

 

Workforce Analytics

Analytics teams allow organizations to harness their data in ways never before imagined, giving companies the power to make more data-informed decisions than ever before.

When we think about business data, we often picture organizations collecting data on their customers to better serve them. However, many organizations are also collecting data on their employees.

Distilling an employee into a series of metrics can dehumanize the management process and make employees feel like nothing more than a number.

Analyzing key employee metrics allows HR departments to better understand their workforce. Tracking employees’ sentiment, productivity, connection to the organization and even diversity metrics can empower HR departments to better allocate resources and improve their workforce’s efficiency. In addition, predictive analytics models can help identify employees at risk of leaving the organization or likely to be promoted.

While workforce analytics may seem like an obvious choice for HR departments, employees may not have the same perspective. Distilling an employee into a series of metrics can dehumanize the management process and make employees feel like nothing more than a number.

More From Mitchell TelatnikHow to Get Started With Social Network Analysis

 

The Future of People Management

Machine learning and AI have promising applications in human resources. However, switching from human to machine-driven people management can cause major issues for employee relations and worker morale within an organization. Do these technologies create a new age in productivity or simply remove the human from HR? Time will tell.

This article was originally published on Towards Data Science. 

Expert Contributors

Built In’s expert contributor network publishes thoughtful, solutions-oriented stories written by innovative tech professionals. It is the tech industry’s definitive destination for sharing compelling, first-person accounts of problem-solving on the road to innovation.

Learn More

Great Companies Need Great People. That's Where We Come In.

Recruit With Us