Human capital management leader Workday, which now supports more than 11,000 customers and 75 million users worldwide, has been steadily shaping its role in the AI era. After debuting on both the Fortune 500 and S&P 500 in 2024, the company launched its Agent System of Record in 2025—a move designed to help enterprises harness and govern the growing wave of AI agents.

Against this backdrop, Joe DosSantos, VP of Enterprise Data and Analytics at Workday, joins Clyde Gillard, North American AI GTM Leader, in the concluding chapter of a three-part series. Together, they reflect on the tangible successes and philosophical future of AI. The conversation spans from productivity-boosting AI assistants to the importance of human-centric strategy in an agentic AI era.

Part one of the interview explored data management challenges, access and security, the shift from generative to agentic AI, and the use of agents in analytics. Part two highlighted AI agents’ current limitations, training agents to admit gaps, Workday’s Agent System of Record as a collaboration model, and the importance of ROI and prioritization to realize AI’s potential.

Squeezing time to insight: AI that coaches, not just informs

DosSantos begins by spotlighting a standout use case from the sales arena: a conversational AI sales assistant integrated with Salesforce. Rather than burying reps in dashboards, the assistant proactively surfaces data about pipeline health, sales stages, and customer account insights — including install base, competitive pressure, and ARR/ACV trends.

“You can say, ‘Tell me about what’s in my pipeline,’ It’ll tell them about the install base and the competitive pressures,” says DosSantos. Layered with reasoning capabilities, the system does not just report; it advises. “You’re getting a coach. It may be directionally correct, but you shouldn’t go to the bank with it,” he notes.

This blend of deterministic truth, wherein the data must be correct, with suggestive interpretation, has transformed how Workday equips sellers.

Automating the quarterly review: From two weeks to two hours

Another major productivity win, says DosSantos, is automating Workday’s internal Quarterly Business Reviews (QBRs). Historically, these were two-week scrambles involving repetitive, formulaic queries and slide deck creation.

Now, he says, “We were able to create an agent that anticipates all of the questions that are asked, pre-runs all of the queries, and prepackages all that,” he explains. Though the final step of automating Google Slides generation remains in progress, the team has already slashed QBR prep time from two weeks to two hours.

“All those people can spend all their time trying to figure out what they’re going to recommend and what insights they can generate,” says DosSantos. In other words, AI frees humans to think, not just fetch.

Let the data speak: Reframing data literacy

As the discussion turns toward data quality, DosSantos makes a bold call to action to stop shielding executives from data.

“People ask questions that the data can’t support, and we’ve hidden the hard work of data from the person asking the question,” he says. True data literacy, he argues, is about creating a feedback loop where leaders can directly evaluate and challenge what they see.

“That’s where we’re headed. It’s trying to get the humans to stop blocking executives and leaders from the data. Stop trying to interpret it for them, give it to them, and let them become part of this. And AI puts a tremendous wind in our sails.”

Shared reality: Aligning data interpretation

Moving forward, DosSantos concurs when Gillard shares his experience with inconsistent CRM data, emphasizing the value of clarity and consensus.

He points out that systems like Google illustrate the importance of consistency. When asked the same question twice, Google typically delivers the same answer, unless the query involves time-sensitive data. DosSantos notes that this predictability highlights a broader truth: “The more things change, the more things stay the same.” For data leaders, this means investing effort in clarifying intent so AI can truly deliver value and improve user experiences.

“If you’re pulling different numbers at different times, you’re sunk,” he states. That is why Workday prioritizes metadata discipline of having date effectiveness, data pull times, and consistent querying logic. “You can train the AI to only answer questions a certain way… So again, these tools are putting wind in our sails, and data governance might be cool again.”

Agentic AI and human strategy

Looking ahead, DosSantos offers a radically different vision of where the world is headed with agentic AI, one that puts humans at the heart of AI orchestration.

“We spend a lot of time trying to make sure it’s not Terminator, to tell you the truth,” he says. He urges organizations to consider a thought experiment: “If I could hire a thousand people tomorrow, what would I have them do?”

That imagined productivity burst is the benchmark for agentic AI. “Hopefully, if we do this right, it will force us to be real thought leaders in our business,” he says. Dashboards won’t make decisions; people will. And that, he argues, is the greatest opportunity AI brings.

The human in the loop: Making strategy a human imperative

For DosSantos, the AI revolution is not about automation replacing humans. It is about intelligent systems enabling better human decisions. He envisions a world where strategy remains deeply human: “How do we make decisions? How do we collaborate? How do we decide what company we want to acquire and which one we want to divest?”

The future of enterprise success lies in the synergy between automated insights and human judgment, he concludes.

CDO Magazine appreciates Joe DosSantos for sharing his insights with our global community.