AI Healthcare Financial Reconciliation Agent

How Cassidy automates this using AI

Step 1: Ingest ERA and bank data

The Workflow triggers when new ERA 835 files arrive via secure transport or clearinghouse connection, while simultaneously pulling EFT deposit data from bank feeds for reassociation.

Step 2: Validate and parse the ERA

Cassidy performs SNIP-style validation—checking envelope integrity (ISA/GS/ST), verifying CARC/RARC codes against national code sets, and confirming segment-level balancing rules per payer companion guides.

Step 3: Match ERA to bank deposit

The Workflow extracts TRN02 from each ERA and matches it to the corresponding ACH reference on the bank statement, confirming that BPR02 equals the deposited amount and queuing discrepancies for exception handling.

Step 4: Post payments and adjustments

Cassidy maps payer and provider identities, then posts CLP claim summaries and SVC service lines to your practice management system—applying CAS adjustments to update patient responsibility, write-offs, and net paid balances.

Step 5: Process PLB adjustments

The Workflow identifies provider-level balance entries (recoupments, interest, withholds) and creates offsetting journal entries, ensuring the final reconciliation balances BPR02 against posted claims plus PLB adjustments.

Step 6: Route exceptions for review

Items that fail validation—balance breaks, TRN mismatches, invalid codes, or zero-pay remittances—are automatically queued with root-cause categorization for your team to remediate.

Step 7: Generate reconciliation reports

Cassidy produces audit-ready reports showing ERA-to-EFT matches, PLB entries, outstanding exceptions, and GL summaries—giving your team complete visibility from bank deposit to posted claim.