AI in Healthcare Jobs: Filling Gaps, Raising Risks
Startups, investors and caregivers show how agentic AI is reshaping the future of work in a system under strain
What’s Inside:
How startups are focused on AI to ease labor shortages in healthcare
POV from investor Sunny Kumar on the economic reality of job displacement
Workforce data and policy shifts accelerating healthcare labor crisis
At a gastrointestinal clinic in the Pacific Northwest, the backlog of patient referrals stretched six months. So four staff members worked full-time to process the paperwork, route patients and handle insurance pre-authorizations.
Then the clinic implemented Lumi, an AI-powered assistant built by Insight Health.
Within weeks, the backlog vanished, and those four employees were reassigned to other work.
For Jaimal Soni, co-founder and CEO of Insight Health, the moment showcased how AI can improve efficiency. He said it was also a sign of what’s on the horizon.
“The real opportunity isn’t just faster documentation,” he said. “It’s offloading the incredible amount of routine clinical work that providers face each day.”
That offloading is happening fast. Insight Health is one of a growing number of startups deploying AI to help plug labor gaps in healthcare, a system that by many accounts is running out of people.
The U.S. is projected to face a shortage of more than 500,000 nurses by 2030, with 193,000 RN jobs opening up annually through 2032, according to RegisteredNursing.org. Meanwhile, STAT News reports a potential deficit of 87,000 primary care physicians, especially in rural communities.
And burnout is compounding the crisis. Turnover rates among nurses now exceed 16%, with many leaving the field due to the workload and chronic stress, according to AAG/Health.
From Backlogs to Agentic Infrastructure
However, with traditional staffing models faltering, AI is moving from pilot to production. Already, 63% of healthcare and life sciences professionals are using AI and 86% believe it’s essential to their organization’s future, according to a 2025 NVIDIA survey.
In fields where the work is repetitive, rules-based or information-dense, AI agents are the new infrastructure, stepping in as healthcare institutions face more work and fewer people to do it.
Insight Health, for instance, focuses on the high-impact pieces of the clinical workflow: patient intake, referral processing and pre-visit history collection. Lumi, the company’s virtual assistant, supports more than 1,500 clinicians. Since launching in 2023, the system has handled more than 120,000 patient interactions.
“We’re not building this to replace clinicians,” Soni told me. “We’re building it because there’s 10 times the work and not enough people to do it.”
But for all the talk of augmentation and filling in the gaps in the workforce, the implications for jobs aren’t as simple. The clinic that used Insight Health to eliminate its referral backlog didn’t lay anyone off. But it also didn’t hire new staff as patient volume increased.
In healthcare, where burnout and attrition are rampant, that may sound a familiar tone. But in the long run, the lines between efficiency and elimination may blur.
For example, look at the work of Resolvd AI, a startup that announced a $1.6 million pre-seed round in August.
Resolvd Co-founder Ananth Manivannan describes the company’s platform as “agentic infrastructure” for healthcare operations. Its agents manage complex hospital workflows, tracking surgical supply costs, validating insurance contracts, updating billing systems and responding to staff emails.
At Resolvd, Manivannan often describes the AI agent as a colleague rather than a tool. One agent monitors surgical supply costs and flags anomalies for the finance team. Another validates insurance contracts, saving hours of back-and-forth emails. Still others help administrators keep track of billing or even summarize staff communications.
“We’re not replacing people,” Manivannan said. “We’re removing work they were never trained to do.”
The point, he argues, isn’t that AI does everything. It’s that each agent takes the “swivel-chair” work off a human’s desk, letting nurses, doctors and managers focus on the parts of their jobs that matter most.
The Investor Lens: Where Jobs Disappear First
That view of AI as a digital co-worker is spreading. But it’s not always accurate.
Sunny Kumar, a partner at Informed Ventures, said the impact of AI on certain healthcare jobs is real. Medical scribes, he noted, have been “virtually eliminated overnight,” replaced by AI-driven documentation tools.
The NVIDIA study confirms that AI tools in healthcare are moving swiftly out of the trials process and they’re undergoing rapid deployment in the market among healthcare professionals and admin officials.
Kumar described three areas where he’s seeing adoption: administrative automation, AI-enabled care and autonomous care. He predicts call center roles will be impacted next, with prior authorization and triage shifting to automation.
The economics, he said, are straightforward.
“If a hospital executive can cut staff by 50% with AI, that’s going to be an option they consider,” he said.
The Risks of Overreliance
However, in his call with me, Kumar framed AI adoption as a response to unavoidable pressures.
And he noted that not every category of healthcare work is under threat. He sees promise in what he calls “AI-enabled care,” tools that help clinicians make better decisions or free them from hours of paperwork.
He also cautioned against over-reliance, pointing to studies where doctors who used AI support systems actually performed worse once the technology was removed.
“These tools can make people better at their jobs,” he said, “but we need to be careful about how they’re deployed.”
And while Kumar points to real displacement in specific categories, workforce data shows that most jobs aren’t going away wholesale. Instead, AI is disrupting tasks within them.
New data from TechWolf’s Workforce Intelligence Index, based on 2 billion job postings across 1,500 companies, found that only 18% of tasks across all sectors are fully automatable. Most (62%) remain fully human, while 38% fall into a gray zone of disruption, augmentation or reskilling.
In healthcare specifically, the Index found that 67% of roles are best served by human–AI collaboration.
That gray zone is where companies like Insight Health, Resolvd and others are operating. The danger, as Kumar points out, is that the line between augmentation and elimination can blur once hospital boards and CFOs start measuring return on investment.
Policy Shocks Accelerate the Labor Crisis
Healthcare policy adds another layer. Digital health founder Jillian Miranda recently posted on LinkedIn of a “perfect storm” taking place where abortion bans, Medicaid cuts and rural hospital closures are accelerating staff shortages. Idaho, for instance, has lost a third of its OB-GYNs since 2022, according to The Journal of the American Medical Association.
Meanwhile, fewer medical residents are applying to train in states with abortion bans in place. Since doctors typically launch a practice where they complete their residencies, access to medical care in certain regions will undoubtedly worsen in the next few years.
These political and structural shocks make it clear: the shortage problem isn’t only about clinicians leaving the field. It also falls on the 53 million unpaid family caregivers, often untrained and overwhelmed, now carrying more of the burden.
That’s the gap Liz Tarullo, founder and CEO of Nolia Health, is addressing.
Her company focuses not on hospitals or clinics, but on the overlooked workforce of family caregivers, who are taking on more responsibility for aging relatives.
These caregivers, she argues, are effectively “first responders in the home,” yet they’re navigating the crisis without training or support.
Nolia Health combines clinical support with a 24/7 AI companion called Noli. The tech-enabled system creates personalized care plans and provides emotional guidance with monthly clinician check-ins.
This kind of support simply didn’t exist before new AI tools and billing codes made it possible.
“We couldn’t have built this company five years ago,” Tarullo said.
For Tarullo, AI and tech solutions aren’t removing jobs but filling a void. Few support systems exist for unpaid family caregivers, she said.
“Our healthcare system is at a breaking point,” Tarullo said. “We don’t have enough clinicians, and we have a population that is aging at an unprecedented rate.”
AI Extends From Clinics to Homes
In the case of Nolia Health, the AI agent doesn’t replace a nurse or social worker. In fact, it replaces no one. It provides guidance where there would otherwise be none.
Whether it’s Insight Health clearing referral backlogs, Resolvd automating hospital operations or Nolia Health guiding family caregivers, healthcare AI is no longer experimental. It’s in the clinics, the call centers and in our homes.
Healthcare systems still depend on humans, but the way work is divided is being rewritten in real time.
That reality underscores why AI technology is arriving not as a choice but as a necessity.
For now, the immediate effect is clear. In the face of a national workforce shortfall, tightening budgets, rampant burnout and changing reimbursement models, AI is showing up in healthcare not just as a productivity tool, but as a structural patch for a strained system.
For more from Liz Tarullo, listen to my conversation with her on The Venture Variety Show, where we go deeper into her founder journey and how AI is reshaping caregiving.