A mid-sized hospital group operating across multiple facilities with accounts at several financial institutions — processing thousands of deposits monthly across check, ACH, collections, and card payments, with posting spread across multiple practice management systems.
The client’s finance team described their reconciliation process bluntly: it was a monthly firefight. Every month-end, a dedicated team pulled bank deposit reports from multiple institutions, downloaded posting data from multiple practice management systems, and spent weeks cross-referencing in Excel. By the time they finished, the financial reports they produced were already stale — and leadership had learned to treat the numbers as directional rather than precise.
The deeper problem was one the finance team had flagged repeatedly: discrepancies between bank deposits and posted payments went undetected for days or weeks. A missing deposit or an unposted remittance might not surface until the month-end reconciliation process — by which time the trail was cold and the resolution window had narrowed.
But the issue that finally triggered the engagement was more fundamental. The client’s existing process matched bank deposits to ERA and EOB files — not to what was actually posted in the PMS. Their controller pointed out that this meant reconciliation had a built-in blind spot: it verified that a remittance was received, but not that it was correctly posted. Posting errors, partial posts, and unreconciled cash were invisible until they showed up as variance in the general ledger.
NDS was brought in to replace the Excel-based reconciliation process — but the first conversation was about data, not software. NDS needed deposit data from all of the client’s bank accounts and actual posting data from their practice management systems. The emphasis on PMS posting data — not ERA files — was the key architectural decision. It meant NDS was reconciling what hit the bank against what was genuinely recorded in the billing system, closing the blind spot the client’s controller had identified.
The system’s fuzzy logic matching engine matched deposit records to posting transactions using key data fields — even when data didn’t align perfectly due to bank-specific formatting, timing gaps between deposit and posting dates, or partial payments split across multiple remittances. Matching rules were configured to the client’s specific workflows and bank account structures.
After automated matching, a GUI presented unmatched deposits for human review. Staff could research and select the best-match posting records for manual reconciliation. Those human decisions were captured and fed back into the matching algorithm — and this learning loop turned out to be more impactful than expected. Within the first 60 days, the automatic match rate improved noticeably as the system absorbed the team’s judgment on edge cases specific to the client’s payer mix.
A centralized reporting dashboard gave cash posters, AR staff, and finance a single source of truth — with real-time visibility into matched, unmatched, and in-progress deposits across every bank and facility.
Pilot with one bank account and one PMS. The initial match rate was 87% — lower than the target. Investigation revealed that one of the client’s banks provided deposit data with truncated reference numbers. NDS added a normalization rule to handle it, and the match rate climbed to 91% within two weeks.
Expanded to all bank accounts and PMS instances. Matching rules configured per facility. Match rate reached 94% across the full dataset.
Daily reconciliation in production. The team that previously spent weeks on month-end reconciliation shifted to exception handling only.
Automatic match rate — but the client noted this mattered less than the fact that the remaining 6% was surfaced immediately for human review
Reduction in month-end close time — from weeks to days. The controller called this the most operationally significant outcome.
Reduction in reconciliation staff effort — two of three dedicated staff redeployed to AR follow-up
Identification of missing deposits — first week in production, the system flagged a $34K deposit credited to the wrong bank account by the payer
Unexpected outcome: The AI surfaced a pattern of modifier misapplication on a specific payer that had gone undetected for over a year. The client's billing team used this to recover underpayments on previously adjudicated claims — a recovery effort that wasn't part of the original project scope.
What the Client Said
"The match rate gets the attention, but what actually changed our operation was knowing every morning exactly what’s matched and what needs work. Before NDS, reconciliation was something we dreaded. Now it’s something we trust."
Discover the way forward through more case studies.
Share your sample bank deposit reports and posting data, and we’ll build a proof of concept showing how NDS matches, reconciles, and surfaces discrepancies — tailored to your bank accounts and practice management systems.
© 2026 NDS InfoServ. All rights reserved.