Orchestrators / Payor / AI Member Retention

01
Section 01

Problem

Members rarely leave a plan. They leave one problem the plan never heard about.

Disenrollment is usually traceable to something specific and fixable: a drug that suddenly costs more, a denial nobody explained, a physician who left the network, an ANOC letter the member couldn’t parse. Each of those is resolvable in a single conversation — the covered alternative, the actual copay, an in-network doctor booked before the call ends.

Switching is also a clinical event. A member who leaves mid-treatment loses the oncologist who knows the case, restarts prior authorizations, and often goes weeks without a medication while the new plan catches up. A member who stays is a member whose care does not restart from zero.

The conversation has to happen inside the ANOC or OEP window, before the decision is made. Today the plan learns the reason from a disenrollment report months later, too late for the member and too late for the next benefit year.

Every 1% of churn is roughly 130 members at $13K in annual revenue.

Reaching every at-risk member inside a several-week window has not been staffable, so plans have managed retention as an analysis problem rather than a conversation.

02
Section 02

Use Cases

  • ANOC
  • Super Physician Change
  • Formulary Alternatives
  • LIS / Extra Help Enrollment
  • OEP Intercept
03
Section 03

Features

Features:

  • Interlinked RAG — reads the benefits policy, cross-references the formulary, and quotes the right copay in one turn.
  • Plan comparison at 99% and benefits verification at 95%.
  • Remaining-deductible (97%), co-pay (97%), and out-of-pocket-max (98%) calculation.
  • Formulary tier lookup for covered alternatives.
  • COB / dual-coverage logic.
  • Prior-authorization vs. pre-certification distinction (97%).
  • EOB explanation at 99%.
  • CMS marketing and communication rule adherence including AEP/OEP restrictions at 92% vs. 61% for the best frontier model.
  • LIS / Extra Help enrollment with auto-form fill.
  • Provider Match supervisor model with in-network physician lookup and in-call booking of a first visit when a member’s doctor leaves the network.
  • Trust Building supervisor model for AI hesitancy.
  • Health Literacy supervisor model.
  • Tangent and Reading Between the Lines supervisor models to catch the real reason behind a stated one.
  • Mid-call language switching.
  • Multi-call Memory.

Potential features:

  • ANOC plain-language walkthroughs.
  • Churn-risk flagging.
04
Section 04

Demo Calls

AI Member Retention Clip 1
AI Member Retention Clip 2