AI Enrollment Reconciliation Agent

How Cassidy automates this using AI

Step 1: Ingest enrollment and billing data

The Workflow triggers on your reconciliation cadence daily for EDI monitoring, monthly for invoice cycles. Cassidy pulls eligibility snapshots from your ben-admin/HRIS, payroll deduction files, COBRA administrator exports, outbound 834s with acknowledgments, and carrier invoices (PDF, CSV, or portal downloads).

Step 2: Normalize and align coverage periods

Cassidy standardizes identifiers (EEID, subscriber ID, SSN), plan codes, tiers, and effective dates across all sources. The system aligns data to your coverage period and applies carrier-specific retro windows and proration rules.

Step 3: Match records and detect variances

Using hierarchical identity matching and fuzzy logic for names and addresses, Cassidy compares eligibility to carrier rosters and invoices. Each record is flagged as exact match, unmatched, or match-with-differences with field-level checks on plan, tier, coverage dates, rates, and ER/EE splits.

Step 4: Categorize exceptions by root cause

Cassidy classifies every variance into actionable buckets: enroll/term in transit, retroactive adjustments, rate or tier mismatches, COBRA misalignments, failed EDI, or eligibility errors. Severity and dollar impact are surfaced for prioritization.

Step 5: Route exceptions and trigger remediation

High-impact discrepancies route to HR, payroll, or benefits teams via Slack or Teams with full context. Cassidy can open carrier tickets, queue corrective 834s, or flag items for manual review while auto-resolving low-risk timing and rounding variances.

Step 6: Generate financial close-out and audit trail

Once variances are resolved, Cassidy produces reconciled invoice summaries targeting zero variance, consolidated payables, GL-ready journal entries, and audit reports with line-level logs of every exception, adjustment, and resolution.