The following is a guest article by Michael Serbinis, Founder and CEO at League
Summary
- Healthcare has always met rising demand by adding people, but that model no longer works, and AI agents that give every member and patient an always-on care team, called the infinite care team, are the way forward.
- Trust is the barrier: 89% of executives call agentic AI very important, only 16% have reached broad deployment, and 99% would delay a launch rather than skip guardrails.
- Start with one use case, such as benefits navigation or scheduling, but build it on a platform where agents work as a team and share the same safeguards, so the next launch doesn’t start from scratch.
- AI Front Door for payers and AI Care Companion for providers run on that platform with five built-in safeguards and fine-tuned decision models for safety-critical calls, so organizations can launch in as little as 12 weeks.
For decades, healthcare has had one answer to rising demand: hire more people. More nurses. More call center seats. More care coordinators. The math simply doesn’t work anymore. Demand keeps climbing, and hiring more staff is no longer feasible. In a recent Healthcare Dive survey of 150 U.S. healthcare executives, 83% cited reducing operational cost as a reason to invest in agentic AI, and 52% cited staffing shortages.
We envision a different future. Imagine that every member or patient has a care team that never sleeps. Agents know them, answer their benefits questions at 2 a.m., find the right care, check in between visits, and bring in a person the moment one is needed. And this isn’t only available for a lucky few. It’s for everyone. We call it the infinite care team, and it’s where healthcare is headed.
Most executives see this future, too: 89% say agentic AI is very important to their strategy and 83% want to use it primarily to improve their customer experience. But deployment is the hard part, with only 16% saying they’ve reached broad deployment. What’s slowing them down is trust. 99% would rather delay a launch by months than go live without full guardrails, and 37% have already slowed down to build in safety guardrails. We believe that healthcare organizations shouldn’t have to pick between moving fast and launching safely, and the organizations that refuse to compromise will be the ones that put an infinite care team in reach of every person on the planet.
Start Where Members And Patients Ask Most
Executives know where they want to start: small. In the Healthcare Dive survey, 75% want to deploy one use case then expand. For most, that first use case is about getting members and patients answers. Benefits navigation leads at 54% and climbs to 59% among payers, while providers lean toward appointment scheduling at 53%. Care finding and chronic condition management also top the list of priorities.
I agree with starting narrow. I disagree with starting over every time. A different vendor, tool and set of safeguards for every use case adds complexity and technical debt with each launch. It also means a member who checks benefits with one agent and finds a specialist with another has to repeat themselves.
A better approach is to start with a single use case, but deploy it on a platform built to support the next one, and the one after that. When agents are orchestrated to work as a team, they can hand off context to each other in the same conversation. That means a member who asks if a specialist visit is covered can immediately book an in-network doctor without explaining themselves again. And because every agent runs on the same built-in safeguards, organizations can launch their tenth agent with the same confidence as their first.
AI Front Door and AI Care Companion
That thinking is behind two solutions we’re launching at HLTH this year: AI Front Door for payers and AI Care Companion for providers.
AI Front Door gives every member one place to get answers and take action. Benefits navigation, care finding, care gap outreach and health coaching agents work as one orchestrated team behind a single entry point, in chat, text or in-app. The goal is to reduce calls to the contact center, drive fewer out-of-network visits, and close open HEDIS and Stars gaps.
AI Care Companion gives every patient a care team that stays with them between visits, at home and after hours. A health system picks a target population or use case, from chronic conditions to post-surgical care, and we configure our solution so that agents can handle scheduling, symptom check-ins, care plan questions, caregiver workflows and escalation to the right on-call team. The goal is better care plan adherence and fewer avoidable ED visits, after-hour calls, and missed follow-ups.
Both are informed by more than 1.8 billion healthcare interactions across the 70 million people League reaches around the world. That history shows us what members and patients ask for most.
Five Safeguards Built In
When executives told Healthcare Dive what would give them confidence to launch AI, three requirements topped the list: built-in healthcare compliance (49%), automatic escalation to a person (48%), and runtime controls on what an agent can say or do (46%).
The League platform has those requirements built in. Every agentic solution that runs on it, including AI Front Door and AI Care Companion, gets the same safeguards from day one, so organizations don’t have to rebuild them for each new agent:
- Approved knowledge only: agents answer from the organization’s provided material, not a model’s general knowledge.
- Guardrails: restricted information is blocked before it reaches the model, and prohibited responses are blocked before they reach a member or patient.
- Eval-gated releases: agents must pass required evaluations against synthetic patient scenarios, edge cases and golden datasets before release.
- Full-context handoffs: escalations carry conversation history and relevant context to the receiving team.
- Continuous oversight: drift detection, revalidation of changes and a per-interaction audit record.
The League platform adheres to HITRUST r2, SOC 2 Type II, ISO 27001, HIPAA and PIPEDA standards and is subject to NIST AI Risk Management Framework governance. With these standards and the five safeguards above already in place, organizations don’t have to spend months building their own. They can launch in as little as 12 weeks, then add agents for more needs on the same platform.
The Era of Healthcare Abundance Has Arrived
For as long as we can remember, healthcare has been scarce for the average person. There have never been enough staff or resources to get every person the care they need at the exact moment they need it. I believe AI will flip that model, from scarcity to abundance. That looks like a member getting an answer to a coverage question in the middle of the night, or a post-op patient getting a check-in the next morning without waiting for a nurse to get to their name on a call list.
When any plan or health system can put safe AI agents in front of its members and patients, everyone can have a care team that never clocks out and brings in a real person when the moment calls for it. Those agents have to earn trust through transparent benchmarks, strict governance, and measurable outcomes. We built AI Front Door for payers and AI Care Companion for providers to that standard.
See it live at HLTH
Seeing these solutions in action is the fastest way to understand them. At HLTH in Las Vegas, Nov. 15 to 18, we’re showing AI Front Door and AI Care Companion live at Booth #2836, alongside the Forge console, where you can watch completions, escalations, and accuracy by use case in real time. Stop by to share the use case you’d start with, and we’ll walk through what it would take to deploy it.
Not going to HLTH? Watch our live AI Care Companion webinar instead to see how it works.
League is a proud sponsor of Healthcare Scene.
About Bob Stanley