What it takes to get your team ready for the agentic era

AI agents can analyze millions of data points in seconds and operate around the clock without fatigue. But convincing one team member to trust that analysis? That can take months.

According to Google Cloud's recent research, organizations achieving ROI from AI agents in under a year prioritize data quality and user adoption above all else. The technology works, but only if people actually use it.

Two leaders from healthcare IT and mortgage lending have learned this lesson firsthand: successful AI agent deployment depends less on algorithms and more on preparing people to work alongside intelligent systems.

The delegation mindset: Reframing AI's role

Jason Bressler, CTO at United Wholesale Mortgage (UWM), hears the same worry from employees constantly: Will AI replace my job? As the technology leader for the nation's number one mortgage company, Bressler knows this fear can paralyze adoption. His message to teams is clear and consistent.

"AI will not replace your job," he tells them. "People who use AI will replace your job."

His role isn't to eliminate jobs (UWM has maintained a no-layoff record for 36 years) but to ensure everyone can leverage AI effectively. The real risk is falling behind colleagues who embrace these tools. For UWM, this means AI literacy extends beyond the IT department to business units, loan officers, and even the independent partners who use UWM's platform.

The word Bressler uses most often when describing AI's role is "delegation." He positions AI as taking over laborious tasks that drain energy and time. "What you need to look at is how it enhances your job and frees you up to take on more complex work," he explained. When a loan officer uses LEO (UWM's loan estimate optimizer agent) to analyze a competitor's offer, the loan officer can focus on the relationship and what the borrower actually needs.

Expert-in-the-loop: Where humans must decide

This delegation philosophy extends across industries. Michelle O'Connor, President and CEO of MEDITECH, sees enormous promise in AI agents. After 37 years in healthcare IT, she also understands the responsibility that comes with deploying them in clinical settings.

O'Connor sets clear boundaries: agents assist, they don't decide. "We don't want to put them down a path of clinical decision. We want to assist them with that," she explained. When MEDITECH deploys agents that summarize a patient's course of stay or nudge physicians to document observed conditions, the technology generates insights and recommendations. The physician makes the call.

Consider a hospitalist taking over care for a patient who's been in the hospital for two weeks. An AI agent can synthesize that entire history in moments. "Summarizing that course of stay gives tremendous time savings and more accuracy," O'Connor noted. "But at the end of the day, the physician will make it their own." The agent does the groundwork. The expert makes the decision.

Aashima Gupta, Google Cloud's healthcare industry lead, calls this approach "expert-in-the-loop," a deliberate evolution of the familiar "human-in-the-loop" concept. The distinction matters: the right expert remains the decision-maker. AI presents information and options, but "the final decision still lies with the clinicians, still lies with the nurses."

O'Connor extends this principle to revenue cycle operations. While AI can suggest billing codes based on documented care, 'we won't make financial decisions on coding patient charts for them. The boundaries are clear: AI assists cleanly and ethically, but authority remains with the expert. In healthcare, O'Connor noted, adoption "will move at the speed of trust."

For a framework on building AI fluency across your organization, see this article.

What leaders must do

Culture change requires leaders who model the behavior, invest in capability, and create space for people to learn.

At UWM, Bressler built an environment where rapid innovation is possible with 2,000 technologists working together. When UWM rolled out LEO, thousands of loan estimates were analyzed within days. That speed only happens when infrastructure and culture are aligned.

“I make sure that not just my IT team knows how to understand the workings of AI agents, but also the other business units and our clients,” Bressler said. This deliberate knowledge-sharing creates a common language across the organization.

O'Connor emphasized that trust must be earned through thoughtful deployment. For MEDITECH, this means being transparent about what AI can and cannot do, setting clear ethical boundaries, and giving clinicians control. Leaders can't simply implement AI agents and hope for adoption. They must demonstrate through action that the technology serves the clinician.

These leaders recognize this is a journey. 'This is the time that I really think healthcare and technology are going to leap ahead,' O'Connor said. Leadership means guiding people through that leap with intention and care.

The human advantage

AI agents can process information faster than any human. They can analyze patterns across millions of data points and work tirelessly around the clock.

They can't build trust with a nervous first-time homebuyer. They can't sense when a patient isn't telling the whole story. They can't take responsibility when things go wrong. The most successful AI transformations liberate and augment human capability rather than replace it. When AI agents handle the routine, repetitive, data-heavy work, humans can focus on relationship-building, creative problem-solving, and ethical reasoning that defines valuable work.

The future belongs to people who embrace AI collaboration, who see agents as tireless assistants. Leaders like Bressler and O'Connor are demonstrating what that future looks like by preparing their people to work alongside these systems, not by implementing technology alone.

The agentic era is here. The organizations that thrive will be the ones with the most prepared people. Learn more about how Gemini Enterprise can help your teams find, understand, and act on enterprise information with AI agents that work alongside your workforce.