Healthcare · Reimbursement Policy Management
F20 Healthcare Payor
Powering an AI-powered reimbursement policy management system to automate policy research, drafting, and maintenance for a F20 health insurer
Impact
Deployed in production
~$50M
Annual savings
<3 Months
Policy update cycle (from 12 months)
5x
Throughput on policy change execution
The challenge
RPM analysts manually research 50+ state regulations, competitor policies, and CMS guidelines to maintain hundreds of reimbursement policies every two years. Each policy requires weeks of research grid creation, alignment summaries, multi-stage peer reviews, and committee approvals. This fragments tribal knowledge across teams and limits throughput.
The solution
The Distillery platform layers AI agents directly into the RPM workflow: auto-generating research grids from competitor data, producing alignment summaries, and drafting V1 policy updates with inline suggestions. Agents learn from analyst accept / reject feedback, continuously improving accuracy while preserving human oversight at every review gate.
AI capabilities
- 01
End-to-end management
Single-platform policy management from a single platform: track progress, kick off new work, and manage every stage
- 02
SME-driven automation
Agent personas encode SME tacit knowledge to auto-draft policies, surface edits, and review artifacts
- 03
Feedback-driven learning
Agents evolve with each accept / reject: storing feedback with context to update prompts and skills over time
- 04
Unified data layer
Links internal tools into a single source of truth with clear provenance across the full policy lifecycle