Workday: Conversational AI will change how we work
In six to 12 months, Workday will roll out generative AI uses for everything from job description writing to contract verification. Conversational AI will be part of the UI.
Ever since the PC revolution, searching for information has remained stagnant. Typically, one would navigate using point-and-click methods and rely on memory to pinpoint the name of a specific report. However, Workday envisions generative AI transforming this process into a fluid conversation between human and machine.
Imagine a manager asking: "How am I doing on hiring?" The machine's response isn't just an answer, but a dialogue, offering varied options and refining its future interactions based on these exchanges enabled by conversational AI.
"You're going to see a lot more of that type of interaction within Workday because it's just how we work as humans," said David Somers, group general manager of products for the office of the chief HR officer at Workday, in an interview.
This dynamic form of conversational AI allows users to adjust report outputs spontaneously. "That's a very different experience than going in and clicking on something" to create a new table, he said.
At this week's Workday Rising conference in San Francisco, the company outlined its initial plans and timeline for generative AI in HR and financial applications. Uses include writing job descriptions, which Workday believes can be done in seconds, rather than hours. Users at the conference seemed interested in what's ahead.
AI is only going to help us get better.
LaShaun FlowersVice president, global HR operations, automation and employee support, Caterpillar
"AI is only going to help us get better," said LaShaun Flowers, vice president of global HR operations, automation and employee support at Caterpillar, who appeared on one of the session panels. "It will make us more efficient. It will make us more effective and reveal things that we didn't even know we needed to be looking for."
Workday's timeline
The initial generative AI capabilities will arrive in six to 12 months. All the major HR vendors are adding generative AI capabilities, with release dates that vary by months. The company doesn't aim to incorporate a conversational AI interface across all its offerings, but it will be a prevalent feature, according to Workday representatives.
Generative AI "is quickly becoming the new automation," Nucleus Research analyst Evelyn McMullen said. "Job description writing is first on the list of use cases for HCM vendors."
Workday's generative AI plan also encompasses features such as reading and summarizing contracts and identifying anomalies, generating content, formulating employee growth plans, offering personalized knowledge resources, and expediting procurement through writing automation.
One capability will be text-to-code, which can use conversational AI and natural language instruction in building an application. The principal value of this will be its ability to act as an "accelerator" for a skilled user to speed up the development process, said Shane Luke, vice president of AI and machine learning at Workday, in an interview. "The machine does not write the code -- it suggests code that the engineer then works with," he said.
For Workday, Luke said the capability will speed up its development process and enable it to ship products faster. Text-to-code could also help HR workers create things such as dashboards without coding skills.
Despite these advancements, Workday emphasizes human involvement in all generative AI functionalities, from crafting job descriptions to content creation and coding.
To avoid the risks associated with large language models that gather data from the public internet, Workday is using "enterprise large language models" that rely on high-quality data and that already have privacy, security and intellectual property protections built in, said Sayan Chakraborty, Workday co-president who heads its product and technology organization, during a presentation. "When we are training models, we know where the data comes from," he said.
Patrick Thibodeau covers HCM and ERP technologies for TechTarget Editorial. He's worked for more than two decades as an enterprise IT reporter.
