At Workday Carl Eschenbach Says It Still Takes People

When Carl Eschenbach stepped in as CEO of Workday in early 2023, AI had just begun reshaping the software industry with unprecedented force. While many legacy companies were scrambling to understand what generative AI meant for their future, Eschenbach, a former Sequoia Capital partner, saw an opportunity to accelerate Workday’s long-term vision: to be the definitive system of record not just for people and their paychecks but for AI agents and the digital labor they perform. “We’ve been managing our clients’ most critical assets—their people and their money—for two decades,” Eschenbach said. “Now, we’re adding digital workers to that mix.” As AI agents become fixtures in the modern workplace, businesses are facing a new operational complexity on how to manage non-human workers with the same level of governance, compliance, and performance tracking as human employees. That’s the challenge Workday’s Agent System of Record (ASR) aims to solve. “We are in the business of work—workplace, workforce, work. These all fall under 'work',” says Kathy Pham , Vice President of AI at Workday. “As we think about agents and AI, we think holistically about everything that makes up the ecosystem.” The ASR platform unifies the management of AI agents and human workers into a single interface. Whether the agents are Workday-native or third-party integrations, ASR functions as a secure command center for onboarding, assigning roles, managing performance, and even offboarding. It’s a bold step toward the operationalization of AI in enterprise environments. “It's human and agent collaboration,” adds Conor McGlynn, senior director of product management at Workday. “It’s really a digital workforce and how that digital workforce can be managed through the Agent System of Record.” The rise of agents has been fast but not frictionless. Enterprises today face a proliferation of tools, mounting costs, and compliance headaches. Workday’s ASR offers a centralized solution to manage agent sprawl, enforce identity and access controls, and monitor usage across systems. “The first thing I always tell folks is to write down what you're solving for,” Pham advises. “Then look out across your tech stack. You might already have an agent that helps.” And many already do. Workday reports that its AI tools have driven a 49% increase in financial planning and analysis efficiency and a 52% acceleration in processing accounts payable. These automations are tailored to reflect the customer’s own policies and processes, thanks to contextual awareness embedded within the platform. Take the Policy Agent , an AI assistant that reads and understands a company’s HR and corporate policies. Instead of HR teams fielding hundreds of repetitive queries, the agent automatically delivers relevant information to managers and employees based on their tasks. “For our customers, the overhead of answering all those questions is huge,” McGlynn explains. “Policy Agent, right in the Workday interface, can instantly answer them—reducing cost by cutting down ticket volume and support load.” Unlike many peers that rushed to monetize copilots with blanket pricing hikes, Eschenbach took a more deliberate approach. “Copilots should be built in, not bolted on,” he argued. Instead, Workday is monetizing AI through a three-part model: Seat-based pricing for features that apply to all users. Role-based agents , such as a recruiting or payroll bot, priced for specific functions. Consumption-based API access via ASR’s gateway for developers and third-party vendors. “We’re not just a software provider,” Eschenbach noted. “We’re becoming the operating system for how companies run their human and digital workforces together.” But for all the power and potential of AI, Workday insists that human oversight must remain at the core. As Chair of Technology & Society at Workday, Taha Yasseri puts it bluntly: “The accountability and responsibility remains with humans,” he said. “Your reference point is always humans—even if we aren’t the best benchmark.” AI models, especially those embedded within enterprise systems, must be carefully evaluated for bias and fairness. Yasseri emphasizes that while machines can assist, they must be aligned with human values and never allowed to surpass them in opacity or ethical ambiguity. “We can make sure the machines are as bad as us—or maybe even better—but definitely not worse,” he said. Workday’s recruiter agent , born from the acquisition of HiredScore, has boosted recruiter productivity by 50% and reduced time-to-hire by up to 40%. Other AI products like Talent Optimization and Talent Mobility are reducing attrition by connecting employees with new internal opportunities, based on their skills. “It’s not about pedigree anymore,” Eschenbach noted. “It’s about skills. And now we can actually match people to work based on that.” This shift from static roles to dynamic talent management reflects Workday’s deeper bet: that AI won’t replace people, but rather en

At Workday Carl Eschenbach Says It Still Takes People

Keep reading. It's free.

Related stories