Workday Inc. announced the creation of Workday AI Research, a specialized technical team designed to improve the reliability, trustworthiness, and efficiency of artificial intelligence within enterprise environments.
Gerrit Kazmaier, president of product and technology at Workday, stated that as AI agents gain the ability to retain context and act on behalf of workers, companies encounter difficult issues regarding privacy, auditability, and accuracy. He noted that standard models often cannot resolve these specific enterprise-grade challenges.
The new research group aims to provide the scientific rigor required to build systems that organizations can deploy at scale with confidence.
The findings generated by Workday AI Research will influence internal AI development while also contributing methods and evaluation practices to the wider academic community.
The team intends to investigate several key questions surrounding current adoption concerns, such as whether multiple AI agents can outperform single agents in decision-making, how to ensure agents remember relevant information without storing excessive data, and the effectiveness of instructions for AI to forget specific details.
In addition to the new research division, Workday introduced the Workday AI Research PhD Fellowship. This program supports doctoral candidates working at the intersection of artificial intelligence and enterprise software. Selected fellows receive an annual unrestricted gift of $50,000 to their universities, direct mentorship from a Workday researcher, and early access to career opportunities within the company.
