How AI is Transforming Workforce Management

Artificial Intelligence (AI) was founded in 1956, long before workforce management (WFM) systems became a staple, and has made advances since its inception. Early models specified all possible choices to enable the computer to make decisions. Today, machine learning (ML) uses flexible models that enable the computer to make choices, including options not explicitly defined.

NICE has made a significant investment into AI and ML. Recent advancements include learning models that find hidden patterns in the historical data used to generate forecasts of volume and work time. NICE WFM also has an AI tool that determines, from a series of more than 40 models, which will produce the best results for each work type being forecasted.

NICE began leveraging AI and ML long before AI became a buzzword. Here’s how:

Skill Use Assessment Intelligence: Determining schedules and efficiency gains when using multi-skilled employees is dependent on the ability to estimate when and to what extent an employee needs to be shared across work streams. The NICE WFM skill usage assessment is based on predictive analysis embedded in the discrete event simulator. Using this approach, no historical data is analyzed, no forced hierarchy of skill consumption is required, and no assumptions or user input is needed. The result is a highly accurate assessment of skill usage, which is the foundation for determining multi-skill efficiencies and optimal schedules.

Skill Use Efficiency Intelligence: One challenge in WFM is understanding the impact of multi-skilled employees on “required lines” – the number of full-time equivalent workers needed to meet service objectives. Most systems rely on a method called Erlang, which has two assumptions that are not valid in today’s modern work center: that all employees share a homogenous skill assignment and that work items queue to a single skill profile. NICE WFM solved this FTE overstatement by adding intelligence directly to the required line calculation. This algorithm provides the intelligence to artificially “deflate” the Erlang-derived FTE to a value that is trustworthy – with no human interaction required.

Closed-Loop Optimization Intelligence: NICE has also invested in artificial learning in the form of “closed-loop schedule optimization.” This application leverages the form of ML in which the machine learns by being fed large amounts of information, with initial decisions “guessed” by the machine. These initial “guesses” are then fine-tuned through a process of comparison to the expected outcome. This is how NICE WFM solves the schedule optimization challenge when faced with several unknowns inherent in an omni-channel environment. NICE WFM does not need to know the exact skill usage estimates or efficiencies before it can start optimizing schedules.

Schedule Fairness Intelligence: Practitioners have recently focused on employee engagement, which has been shown to boost productivity and performance. NICE has made significant inroads with its Employee Engagement Manager solution, which has multiple capabilities designed to include employees in the complex process of managing schedules to meet customer needs.

In addition, multiple machine algorithms are designed to create a fair workplace that replaces or enhances traditional seniority-based assignment processes. Some examples of this in action include:

  • Adaptive Assignment: When NICE WFM is integrated with NICE Performance Management’s adaptive intelligence capabilities (NICE AWFO), work schedules can be assigned using uniquely identifiable metrics, attributes and preferences of each employee.
  • Preference Persona Assignment: NICE WFM offers a robust persona that each employee self-manages. The capabilities include self-identified work time availability and a customized preference persona. The machine constantly monitors for changes in preference personas and adjusts assignments accordingly.
  • Policy Assignment: NICE WFM has intelligent policies that manage schedule creation and assignment fairly while meeting customer needs. These algorithms are designed to balance the needs of the business and employees while eliminating human intervention.
  • Fairness Assignment: Some people want to volunteer to work certain days of the week, weekends or holidays, while others want to be rotated through the assignments fairly. NICE WFM fairness intelligence monitors assignment history, fairness credits (which can be tied to NICE AWFO), volunteerism, work rules and business need to manage the assignments fairly.

AI’s utility and promise are nothing new for NICE, which has continuously invested in AI and ML to help omni-channel contact centers, back-office operations and branch environments benefit from “the science of getting computers to act without being explicitly programmed.” With NICE WFM, the machine takes on the tasks of learning the uniqueness of each environment and applying intelligence that exceeds the human capacity to process. In doing so, it frees humans to focus on those activities and thought processes that require the human touch.

For more information about NICE, please visit our website at www.nice.com.