UKG is developing artificial intelligence tools with a strong focus on frontline workers, including voice-first systems that integrate with Siri and Google Assistant to help employers manage labour shortages and reduce employee turnover.
The workforce operating platform company is building what it calls “assistive AI” — tools that let workers handle tasks like requesting time off or checking schedules through conversational voice commands rather than logging into traditional software applications, according to Jay Henderson, group vice president of emerging technologies, and Corey Spencer, general manager and group vice-president of AI at UKG
“They’re (often) running to some other situation, and whenever they’re engaging with our technology, it has to fit into their life,” said Spencer during a conversation with HR News Canada. He noted that frontline workers like line cooks, stock clerks, and hotel housekeepers need mobile-first solutions because many lack corporate email addresses or regular computer access.
The company is addressing what Henderson described as mounting pressure on employers to “do more with less or more with the same,” leading to widespread employee burnout. UKG’s clients report three main challenges: accelerating hiring processes, decreasing employee churn, and improving operational efficiency between hiring and turnover, according to Spencer.
Two approaches to AI deployment
UKG is pursuing two distinct AI strategies, Spencer said. The first involves assistive tools for immediate tasks. “I need to book next Tuesday off. I shouldn’t have to open my phone and get to a calendar,” he said, describing how a nurse could simply ask Siri to request time off through UKG’s system.
The system can also help managers handle unexpected absences. “Todd didn’t show up to work. Who’s the next best person to take Todd’s position?” Spencer said as an example. “The AI can check and say, independent of your institutional knowledge, say, Dan is the next best person. Do you want me to assign that shift to him?”
The second approach involves what Spencer called “truly agentic work” — AI systems that function as digital delegates constantly monitoring multiple data inputs including weather, staffing levels, and patient outcomes.
“It’s taking all of that data and it’s working as like a little angel on my shoulder,” he said, suggesting adjustments to managers in real time.
Spencer said the assistive AI tools are pretty much ready for deployment, while the more complex agent systems that coordinate between multiple companies’ AI platforms remain in refinement. He noted that companies including ServiceNow, Salesforce, and UKG are investing heavily to enable their AI systems to communicate with each other.
Voice commands gain traction
UKG has built a voice-first interface, including a Siri integration that allows workers to handle many tasks through voice commands.
“When they try something more complex, Siri can pass them into the appropriate part of the application,” Henderson said.
“Given the audience that we serve, it’s a pretty natural place for them to want to be able to interact, and a mode of interaction that really works for them,” he added.
He acknowledged it took time to discover how important voice interaction would be for frontline workers whose hands are busy throughout their shifts.
The company’s workforce intelligence layer, announced at its Aspire conference, consolidates data from recruitment through scheduling and payroll. Spencer said this semantic data layer captures the complete life cycle of workforce interactions, with about 95 percent of workforce data proving common across organizations and only 5 to 10 percent requiring customization.
Conversational reporting changes data access
The company launched conversational reporting, which lets managers query workforce data using natural language. Spencer said users can ask questions like “what is the average tenure of my employees in the store” and “show me the average hours worked by gender.”
Spencer described a case where a vice president of HR at a major financial institution demonstrated the tool to the executive team. “The CEO was like, ‘How the hell did you get this data? Why don’t I have access to this data?'” Spencer said. The CEO mandated company-wide access within weeks, declaring that no manager could claim they lacked workforce information when they could simply ask the chatbot.
Reception from HR departments
Spencer acknowledged that AI adoption varies significantly among HR departments, which he said are generally “not known for a software industry being on the bleeding edge.” Some organizations have rejected the technology entirely, while others have embraced it.
One HR administrator in their early 50s asked Spencer whether adopting AI meant becoming “the AI administrator for HR,” expressing concern about answering technical questions without adequate experience. Spencer said UKG is developing new training and support programs specifically for administrators, acknowledging the company initially “kind of” treated AI as just another feature.
The company is also testing AI-powered payroll automation. Spencer said beta testers initially told UKG that “over my dead body, will you automate anything in my job” because payroll is “a zero sum game” requiring human oversight at every step. Within six weeks, those same administrators asked whether UKG could reduce notification frequency because they had gained confidence in the system and wanted to focus on higher-value work, Spencer said.
Henderson said customers have standard concerns about security, governance, privacy, and whether their data trains AI models. “There certainly are organizations who are just saying, ‘No, this is not for us. We don’t want things to get better with AI,'” he said.
But many customers recognize the potential for improving employee experience, Henderson said. “Do you know, focus on doing their job more and not using somebody’s software to manage things,” he said, describing how AI could reduce administrative burden.
Future vision for workforce technology
Spencer predicted that within a few years, the way workers interact with UKG’s software will become less dependent on traditional application interfaces. Workers won’t need to “stop your day and log into an application and find the navigation and click on the thing,” he said.
“You won’t have to learn our software, because our software will learn your day of work,” Spencer said, describing a future where UKG meets workers through whatever device or interface they’re already using.
He compared the current AI moment to the early Internet rather than blockchain technology. “It feels a lot more like the early days of the Internet than the early days of the blockchain,” Henderson said, noting that with blockchain he never understood whether it would become practically useful, while AI clearly represents major disruption.
