Workforce management solutions help organizations track and optimize their labor costs while also having a significant impact on employee experience as they are solutions that employees and managers/supervisors use every day. Generative AI has received a lot of hype due to the consumer success of tools like ChatGPT and Bard. Organizations are looking at opportunities to leverage these capabilities. They want to know what internal use cases could benefit most from generative AI. In workforce management, there are a number of use cases where generative AI can have a significant impact including improving digital assistant functionality and delivering greater insights on labor costs and optimization opportunities.
What is generative AI?
Generative AI refers to a system that can create new content, including text, audio and images based on a large amount of existing data. It works by collecting examples and identifying patterns from a vast amount of data and then using those findings to create something new. In the consumer world, for example, generative AI could create a children’s book based on other children’s books by famous authors or create a new hit song using music from top recording artists.
AI is not new to workforce management. Creating forecasts and developing optimized labor schedules have been use cases for AI and machine learning (ML) for several years. What is new is the ability to leverage generative AI to expand the usage and impact of AI for workforce management.
Digital assistants for workforce management
Digital assistants use artificial intelligence (AI), natural language processing and ML to provide a personalized, conversational experience. They can answer questions, make recommendations and even initiate conversations. In workforce management, one of the opportunities to leverage generative AI is to improve the digital assistant conversational experience by making those conversations even more relevant and to provide more assistance without human intervention.
For deskless workers, digital assistants are a valuable resource for getting questions answered in order to take the actions needed. Before generative AI, if a deskless worker had a question about paid time off (PTO) or leave of absence policy, they might ask a digital assistant, but the answer would typically be a link to a policy document or knowledge article. The worker would still need to figure out the answer based on reading the document (or asking HR if they cannot figure out the answer). With generative AI, the digital assistant can provide an actual answer without the worker having to even read the document. If the worker has a follow up question, the digital assistant can continue the dialogue with the context of the previous question and answer. This is similar to how the worker might interact with HR to get the question answered.
Generative AI can support this kind of conversation because the underlying Large Language Model (LLM) can be trained on policy information so it generates answers instead of just links to documents (though it can also link to the source document in some cases, if necessary).
Vendors like Lucy.ai and Leena.ai provide general solutions that can provide answers for all types of queries. Major vendors like Oracle, SAP, ServiceNow, UKG and Workday are adding generative AI to their platforms to enable this kind of interaction.
Greater workforce insights
One advantage of AI, and generative AI, in particular, is the ability to process large amounts of structured and unstructured data to find patterns or make recommendations. In workforce management solutions, very granular time data can be captured and used to help understand labor costs and identify opportunities for cost savings. What can be challenging, though, is marrying this data with the other structured and unstructured data that may exist in the enterprise that may provide additional context and insights. Usually, this involves extracting data from a variety of different systems into a data warehouse or data lake and using business intelligence (BI) tools to find new insights.
Generative AI has the potential to take things one step further by bringing more unstructured data into play. In addition, generative AI has the ability to identify patterns that may not be obvious in both the structured and unstructured data, even to an experienced analyst. Generative AI also explains those patterns in natural language making it easier for managers and executives to understand the insights generated. Finally, generative AI can make recommendations of actions to take based on the insights. A good example of what is possible would be to link questions about (or viewing of) time-off policies to actual absence behavior (missed shifts or other unexcused absences). The system might identify absence patterns (like higher missed shifts on Mondays) via conventional workforce management analytics but making the connection to policy information access (to see what is allowed or not allowed) may not have been an obvious avenue to investigate. Generative AI and its supporting models can surface nonobvious labor cost and workforce optimization insights.
Vendors like Visier and CrunchHR are adding generative AI capabilities to their people solutions. Also, major vendors like Oracle, SAP, UKG and Workday are adding these capabilities to their analytics offerings.
Conclusion
Generative AI using LLMs can be used to improve employee experience and to improve workforce management insights. Enterprise clients should be actively engaging with their workforce management vendors to understand their current capabilities and roadmap to leverage generative AI for these use cases. In addition, if your current workforce management vendor is not well-positioned to provide generative AI capabilities for needed use cases, consider specialist solutions in areas like digital assistants and people analytics to fill in gaps (or to provide a more holistic solution across multiple domains).
Workforce management vendors should make sure that they understand how they will leverage generative AI in their software platforms and make sure that key use cases that leverage these capabilities are on their roadmap.