AccAnalysisAccAnalysis
Revenue Cycle & Financial Operations

Which accounts to work today, scored nightly and re-learned weekly

A recovery operation was prioritising thousands of accounts using a generic credit-rating signal that was right about six times in ten. We built a self-learning model scored on the operation's own outcomes.

At a glance
Status
Delivered
Region
United States
Engagement model
Dedicated team

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 healthcare debt-recovery operation with thousands of active accounts and a finite number of agent hours.

Which accounts get worked, in which order, is the single largest determinant of how much is recovered — and it was being decided using a bought-in credit signal that had nothing to do with this operation's own history.

The brief

The problem

  • Prioritisation rested on an external rating not built for this population.
  • Effort was being spent on accounts with a low likelihood of paying, which is expensive twice over: the cost of the call, and the higher-likelihood account not called.
  • There was no feedback loop — outcomes were not being fed back into the next day's priorities.
  • The measure of accuracy was not being tracked at all, so there was no baseline to improve on.
The work

What we built

An ensemble model

Combining several gradient-boosting and tree-based approaches, scoring every account daily with a probability of payment.

An automated pipeline

Running the scoring on a daily schedule against current data.

A weekly retraining loop

So outcomes feed back into the model and prioritisation adapts as the population changes.

Method

How we delivered it

  1. 1

    Establish the baseline first

    Measure how accurate the existing credit-rating approach actually was, because without that number an improvement cannot be claimed.

  2. 2

    Feature engineering

    On the operation's own history, with the data team.

  3. 3

    Model development and comparison

    Holding out a period for honest evaluation.

  4. 4

    Pipeline

    Daily scoring, monitored, with the score written where agents work.

  5. 5

    Feedback loop

    Weekly retraining on outcomes, with accuracy tracked over time.

Sequence

How it was phased

PhaseDurationWhat happens
1Baseline measurement
2 wks

Accuracy of the existing approach, evaluation method agreed

2Feature engineering
3–4 wks

Signals from the operation's own history

3Model development
4–5 wks

Ensemble build, held-out evaluation

4Pipeline
2–3 wks

Daily scoring, monitoring, delivery into the agent workflow

5Feedback loop
2 wks

Weekly retraining, accuracy tracking

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

Hand-over

What the client keeps

  • Model code and training pipeline in their own repository
  • The feature definitions
  • Held-out evaluation results
  • The baseline measurement
  • Retraining configuration
Stack
PythonXGBoostLightGBMRandom ForestCatBoostSQLScheduled pipeline
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.