This content is part of the Essential Guide: It's a millennial workforce -- here's what HR pros need to know

How AI and machine learning help in upskilling employees

The rapidly changing technology landscape requires companies to change their approach to L&D and deliver more individualized learning. Here's how AI and machine learning may help.

"If your company isn't focused on upskilling, your workforce won't be well-equipped to adapt to new challenges and make the most of new opportunities," according to Jim Link, chief HR officer of staffing service Randstad North America. "Ultimately, your company will lose out on opportunities to innovate."

Indeed, upskilling employees -- that is, teaching staff additional skills -- is becoming an increasingly critical practice for organizations and their HR departments as more industries face skill gaps and hiring markets become more competitive. Companies can take a variety of approaches to upskilling employees. On the technology front, AI and machine learning (a type of AI) promise to enable upskilling programs that help people adapt more quickly to an ever-changing workplace and world of work faster.

Trish Uhl, founder of Owl's Ledge, a learning and development consultancy, said machine learning is being used in a number of ways, including improving instructional content creation, providing visual and auditory feedback, developing better recommendation engines, enabling faster instruction design and personalizing performance coaching.

"The advantage of cognitive technology is that it can handle vast amounts of structured and unstructured data with a speed and accuracy to find training courses that match [employees'] current career experience," said James Cook, global partner of workforce development at IBM. "We believe there is an unprecedented opportunity for HR to drive digital transformation for many industries."

We believe there is an unprecedented opportunity for HR to drive digital transformation for many industries.
James CookGlobal partner of workforce development, IBM

The use of machine learning in upskilling employees is promising but is still in its early phases. It's important to identify useful metrics that can guide upskilling efforts. Third-party tools can make it easier to experiment sooner. In the long run, machine learning could also improve efforts to help employees improve their soft skills, in addition to technical ones.

Early days of upskilling employees with AI

The use of AI as part of upskilling initiatives is relatively new, said Bruce Cronquist, learning and development program manager at Dell. Until six months ago, the company was using only simple analytics to analyze and guide Dell's upskilling program. Most recently, it has adopted the EdCast platform, which uses AI to suggest training paths for employees.

AI and machine learning enable the system to make recommendations beyond a particular topic to analyze what their training teammates and people with similar interests are taking. Cronquist said it is still early to quantify the success of this program, but other companies he has talked with have reported positive results.

Metrics needed for success in upskilling employees

An important aspect of building a better upskilling platform lies in identifying metrics that matter. Although HR professionals may start with their own ideas, it's important to ask other stakeholders. Cronquist said, "Once I have my list, then I determine which metrics to collect to help me hopefully show improvements in what matters to them."

If these metrics are not readily available, then it is important to find an easy way to collect them. For example, when Cronquist was tasked with improving onboarding, he had to figure out what exactly engineering leadership meant. It could have referred to greater satisfaction or a faster process. He used a precision question-and-answer process to determine what software development managers wanted from new hires to become productive sooner.

This data did not exist in any of Dell's existing reports, so Cronquist started collecting metrics from the beginning of projects through to submitting working code into production applications. This was captured for a year before and after experimenting with a better upskilling process. Using these metrics as a guide, Dell cut the time from onboarding to working with the development team's shared code database from 50 to eight days and time to first committing live code into production from 100 to 37 days.

Commercial machine learning tools for upskilling experimentation

A company can either create its own tool or purchase third-party software to bring in AI and machine learning for upskilling employees. Dell decided to engage third-party vendors to experiment faster. For example, Cronquist and the company are experimenting with other software that uses machine learning to identify gaps in the skill sets of educational services employees who want to move up.

Cronquist said, "This is turning into a big project since we are in the process of identifying the skills necessary for each level of each job, then creating some type of assessment to verify those skills." Sometimes, filling in the gaps is as easy as recommending employees take a specific training in a third-party tool on a website. But it's not usually that easy. Dell has started by rolling this out for the education services team, and once it works out the bugs, Cronquist has plans to address the rest of the company.

Companies like IBM are moving to adapt their commercial AI tools for use in upskilling employees. For example, IBM Watson Career Coach is meant to foster employee engagement with cognitive career guidance. Career Coach makes recommendations to each employee for each next career step. The recommendations include multiple job role choices that are likely to suit the employee, with ratings for each based on organizational demand and skill fit. Career Coach prepares employees for their specific career growth path, with relevant upskilling suggestions. Users can modify, at their own pace, their selected career plans, as needs or desires change.

New approaches to upskilling employees' interpersonal savvy

Randstad's Link said, "While AI is useful for many tasks, it will never be able to replicate human creativity, emotional intelligence or complex communication." He believes that efforts around upskilling employees should focus on skills AI doesn't possess, like strategic and abstract thinking, leadership, problem-solving and communication skills. "Interestingly, those are also the types of skills younger employees are most eager to sharpen," he said.

AI can't replace soft skills but may help in boosting them.

According to Uhl, AI and machine learning show promise in making it easier for HR professionals to connect the dots between metrics around challenges like decreasing absenteeism and improving employee engagement. The key lies in using workplace analytics tools for measuring these soft skills and then using AI and machine learning to identify which training approaches support people in becoming measurably better all-around contributors.

Six ways machine learning can improve upskilling

Uhl believes machine learning will increasingly change how corporate training is conducted with improvements to upskilling employees. Here are six ways machine learning can boost employee learning:

  • Personalization: Machine learning is intended to customize the learning and development experience -- moving from addressing needs in a general way to addressing the needs of each individual, where each person receives ongoing contextualized, individualized, personalized skills and capability development over time.
  • Acceleration: With machine learning, the ability of employees to move from novice to expert will be accelerated -- for example, collapsing the development timeline from years to months.
  • Optimization: Feedback loops and ongoing, automated performance support provide reinforcement and practice that can extend skills development.
  • Prediction and prescription: Leading indicators can be used to improve the probability of success by making corrections going forward, rather than by looking back and using historical lessons that may not apply.
  • Efficiency and effectiveness: Machine learning is intended to enable the ability to track real-time program effectiveness and ROI, thereby conserving resources and reducing training investment costs.
  • Impact: Upskilling employees through machine learning platforms is intended to improve targeted performance outcomes, both in terms of inspiring measurable positive people impact and driving demonstrable value creation for the business.

Dig Deeper on Talent management

Business Analytics
Content Management
and ESG