AccAnalysisAccAnalysis
Revenue Cycle & Financial Operations

Six kinds of discrepancy, caught on a schedule instead of eventually

Payment differences between hospitals and the collection operation went undetected until someone reconciled by hand. We built automated reconciliation flows for twelve client facilities, running on their own cadences.

At a glance
Status
Delivered
Region
United States
Engagement model
Shared specialists

Status: Delivered. Described from our own delivery records. Client identity, brand and commercial terms are withheld. Figures, where given, cover the period stated and nothing beyond it.

The organisation

Context

A revenue-cycle operation whose payment records must agree with those held by the hospital facilities it serves.

Where they disagree, money is either not collected, not credited, or credited twice — and the disagreement does not announce itself.

The brief

The problem

  • Discrepancies were found only when someone reconciled manually, which was slow and error-prone in itself.
  • Different facilities operated on different cadences, so one reconciliation routine did not fit them all.
  • By the time a difference surfaced, the window to resolve it had often narrowed.
  • The revenue effect was real but unquantified, because nobody was counting what was being missed.
The work

What we built

Automated reconciliation flows for twelve client facilities

Each running on the cadence appropriate to that facility rather than a single group schedule.

Six categories of discrepancy detected automatically

Each raised with the records that produced it rather than as a bare exception count.

Scheduled distribution of findings

To the stakeholders responsible for resolving them.

Method

How we delivered it

  1. 1

    Define what a discrepancy is

    Six categories, agreed with finance and operations before any code was written. This is the step that determines whether the output is trusted.

  2. 2

    Build per-facility flows

    Since the data shape and cadence differ.

  3. 3

    Run silently for a period

    Comparing automated findings against manual reconciliation.

  4. 4

    Turn on distribution

    Once the false-positive rate was acceptable to the people who would receive it.

  5. 5

    Add new facilities to the pattern

    As they onboard.

Sequence

How it was phased

PhaseDurationWhat happens
1Discrepancy definition
2 wks

Six categories agreed with finance and operations

2First facility flows
3–4 wks

Pattern established on two or three facilities

3Roll-out to remaining facilities
4–6 wks

Per-facility cadence and data shape

4Silent run & tuning
2–3 wks

Automated findings compared against manual

5Distribution & hand-over
1–2 wks

Stakeholder routing, runbook

Indicative phasing for work of this shape. Actual duration varies with data quality, access and decision speed.

Hand-over

What the client keeps

  • The reconciliation flows and their source
  • The discrepancy definitions as documentation
  • Per-facility configuration
  • The distribution routing
Stack
PythonPower Automate DesktopSQLScheduled orchestration
Keep reading

Related

Related case studies

Tell us what your current system can't do, and we'll tell you what it would take to change that.