AccAnalysisAccAnalysis
Healthcare & Social Care

Weekly raw data, turned around without anyone touching it

Hospital insurance data arrived weekly in raw form and needed manual extraction, analysis and charting before clinical and administrative staff could use it. We automated the pipeline end to end and published the results to a portal.

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 healthcare operation receiving weekly insurance data files covering coverage and claims for hospital patients.

Before hospital personnel could act on any of it, someone had to load the files, clean and transform them, run the analysis and produce the charts.

The brief

The problem

  • Every weekly cycle repeated the same manual extraction and transformation work.
  • The delay between the data arriving and the data being usable ate into the week it described.
  • Manual handling introduced variation, so week-to-week comparisons were not always sound.
  • Hospital staff had no direct access — they received output, and had to ask for anything else.
The work

What we built

An automated weekly ingestion and transformation pipeline

Triggered on arrival rather than run by hand.

An analytics processing stage

Producing the measures consistently every cycle.

A web portal with multiple dashboard tabs

For hospital personnel, so the output is somewhere they can go rather than something they wait to receive.

Method

How we delivered it

  1. 1

    Document the manual process

    Exactly as performed, including the judgement calls, before automating any of it.

  2. 2

    Automate ingestion

    With validation on arrival and alerting on missing or malformed files.

  3. 3

    Automate transformation and analysis

    With the same measures every cycle.

  4. 4

    Publish

    To the portal, with the tab structure driven by what staff actually needed to see.

  5. 5

    Run in parallel

    For several cycles, comparing automated output against the manual result before switching over.

Sequence

How it was phased

PhaseDurationWhat happens
1Process documentation
1–2 wks

The manual cycle, captured as it is performed

2Ingestion automation
2–3 wks

Scheduled load, validation, alerting

3Transformation & analytics
3–4 wks

ETL and the measures, made repeatable

4Portal
3–4 wks

Dashboard tabs for hospital personnel

5Parallel cycles
3–4 wks

Automated output compared against manual, then switch-over

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

Hand-over

What the client keeps

  • Pipeline code in their own repository
  • The transformation logic documented
  • Validation rules
  • The portal and its source
  • The parallel-run comparison record
Stack
PythonETL pipelinesSQLWeb portalScheduled 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.