Zachariah Hennessey, Chief Strategy Officer and Executive Vice President of WholeYouNYC at Public Health Solutions (PHS), and Luz Guardado, Senior Health Insurance Navigator at PHS, and a New York State certified navigator, joined Blooming Health to discuss how their organization has scaled Medicaid outreach to more than 600,000 New Yorkers.

They achieved this not by automating people out of the process, but by being precise about which parts of the work belong to technology and which parts never should.

What AI Should Automate

For Public Health Solutions, automation starts with the administrative weight that keeps people from reaching a navigator in the first place.

As Hennessey put it, the goal is "utilizing the technology to take that first step of bringing certain cohorts of people to us" and to "deploy the technology to make the partnerships and coordination around the people more efficient."

That includes screening. Rather than routing every person through the same lengthy process, Public Health Solutions uses logic that adapts based on how someone responds — routing them to the right resource, or redirecting them elsewhere, without making a person wait "days or weeks" for someone to sort it out manually.

Hennessey explains, "You really are being customer friendly and customer service oriented by not taking people through a lengthy in-person process or a lengthy manual process when you could sort of automate the right questions first."

The payoff isn't just speed. It's where the savings get reinvested. Hennessey pointed to the deeper problem in the system: "The connection of people to [care] is so inefficient because of the complexity that we have these runaway administrative costs, and the vast majority of resources are not actually spent on the care." Automating the routine work is how that math starts to change.

What Navigators Should Always Own

None of that automation replaces Luz Guardado's role, it protects it. Guardado described a caseload built on trust, follow-through, and a level of hands-on problem-solving no algorithm can replicate: helping clients with medical bills, calling hospitals, applying for financial assistance, explaining the difference between Medicare and Medicaid, even helping select a primary care provider.

"My role doesn't end after client enrolls and covers," she said, "It's important that the human communication and technology actually bond together."

Hennessey framed the division of labor plainly: the goal is to "cut out all the time and effort involved in getting people to Luz" — so that "when they get to Luz, they belong with Luz, and she can really help them the most effectively." 

Technology's job is to clear the path. The relationship at the end of it is still human.

How AI Helps Prioritize Cases

Scaling outreach doesn't mean treating every case the same. Public Health Solutions layers Medicaid data — lapsed coverage, address changes — with clinical context like pregnancy, chronic conditions, or disability status. That combination lets the organization, in Hennessey's words, "prioritize and triage and lift up for our navigation workforce the most intelligent next right steps for them to take."

It's a model built around a simple principle: not everyone needs the same intervention at the same time, but everyone needs to be seen. As Hennessey said of the deployment question, "who is deploying" a technology matters more than what the technology can technically do — and trust, once established, is what makes automated outreach land instead of getting flagged as spam.

Empathy is Key

Guardado's account of one client, referred for a transplant, denied by her hospital because of a coverage mismatch, is the clearest argument in the entire conversation for why navigators can't be automated away.

She coordinated between the health plan, a social worker, the dialysis center, and the hospital to get the coverage switched. Six months later, her client received her transplant. "She feels healthy and enjoys her life, and she is super grateful to all my effort that I helped her in this process," Guardado said.

That's the outcome technology exists to protect, not to produce. As Hennessey summarized it, the deciding factor in every technology decision Public Health Solutions makes comes back to one thing: "the best interests of the member, the best interests of the person." AI can find people faster, prioritize them more intelligently, and clear away the paperwork between them and care. But the trust that makes a person pick up the phone, and the judgment to walk them through a life-or-death coverage decision, still belongs to a human being.

That's the model worth building toward — not AI instead of navigators, but AI that gives navigators more time to do the part of the job only they can do.

Want to see what a care enablement workflow looks like in practice?

Blooming Health helps organizations like Public Health Solutions reach more Medicaid members, prioritize outreach based on real risk factors, and free up navigators for the conversations that need a human. See how Blooming Health AI-powered Care Enablement Workflows work.