Andrea Rodriguez

Epidemiological informatics · risk · reporting systems design · data visualisation and AI-driven automation

Using your data to drive process change

For boards and executives of large health, community care, aged care and government organisations. See the work, remove what adds nothing, prove the change.

Start a conversation

Andrea Rodriguez · Sydney

Selected work

Three ways of showing change to a board

Illustrative data

Where each organisation sits, year by year

Every organisation, every year, against the group.

Each dot is one organisation; a larger dot means more occupied bed days. Gold mark: pooled rate for all twenty-four organisations that year. Outlined dots: the highest and lowest organisation in 2025, named.

Illustrative data

Separating a real signal from ordinary variation

One month that matters; the rest is ordinary variation.

Monthly rate per 100 admissions. Centre line: the average over the twenty-four months. Shaded band: between the lower and upper control limits. Gold ring: a month outside the control limits.

Illustrative data

A process before and after redesign

Nineteen steps become twelve.

Mauve markers: the twelve steps kept, placed by elapsed day. Pink markers, shown as the chart plays: the seven steps removed. Dotted line: the days taken out of the process.

Services

Three disciplines, applied together

Process redesign

Lean and Six Sigma applied to the way work and information move between teams, services and systems.

  • Value stream maps of the current and future state
  • Cycle time, lead time and waiting time measured at each step, so delay has a number
  • Duplicated entry, hand-offs and rework counted, and the constraint found

Reporting systems design

Measures a board, an executive committee or a regulator can rely on, built on clinical epidemiology.

  • Every measure defined and agreed before data is collected
  • Statistical process control, so real change stands apart from ordinary variation
  • Risk read from results, within your own risk framework

Data visualisation and automation

Reports that can be read in minutes, built so they assemble themselves.

  • Each chart with its data and a plain-language reading
  • Board papers in print that match what is on screen
  • Hand-compiled reports replaced by one calculation behind them all
  • Screens designed around each user's tasks, with every click counted and tested

A place to start: process review and recommendations

A fixed-scope review of one process, over two to four weeks. The process is followed end to end, cycle time and waiting time are measured, the data and reports behind it are checked, and risks are noted. You receive a short report of recommendations in priority order, each with the gain it should bring.

What you receive

Work your organisation can use the day it is handed over

  • Current and future state mapsLead time, cycle time and waiting time measured for every step
  • A measure set your board can rely onEach measure defined, sourced and placed on a control chart
  • A working prototypeThe new process and its reports, tested with the people who will use them
  • A tested specificationReady for your technical team or builder to take forward

Reports that build themselves

Board and executive reports Regulator and funder returns Comparative reports Print matching screen

Grounded in

Master of Quality in Healthcare, University of Newcastle Clinical epidemiology Lean Six Sigma Statistical process control Applied AI and automation User experience design

Every change is shown before and after, in your own figures.

Organisations worked with

  • Ramsay Health Care
  • Joint Health Command
    Department of Defence
  • The Whiddon Group
  • New South Wales Health
  • Victorian public health

How the work runs

From the way it is to the way it could be

1

See the current state

Follow the work end to end.

2

Count what adds no value

Re-entry, waiting, hand-built reports.

3

Design the future state

Capture once; one calculation.

4

Build and test a working version

Short cycles with real users.

5

Hand over the evidence

A tested specification.

A recent engagement

Certification and reporting, redesigned for an industry certification body

19 to 12steps in the process
7 to 1systems the information passes through
6 to 0points where staff enter the same information again
6 to 1types of report put together by hand

105 build cycles in one month, tested before any platform spend. Client kept confidential.

Working prototypes

See the new process working before you buy it

Built with AI in weeks. Method, statistics and sign-off stay with experienced people.

  • Every design decision recorded, with its reason, before any code is written.
  • Every calculation checked by a test, and every screen tested as each user would see it.
  • Synthetic data throughout design; your records stay inside your organisation.
  • Nothing moves on without your sign-off.

Fractional AI lead

An AI lead for your organisation, a few days a month: where AI belongs in your processes and reporting, how it is governed, tested and kept secure, and what it should cost.

Technical architecture and infrastructure

Platform design, security, hosting and integration through an experienced technical architect partner.

Andrea Rodriguez

Contact

Tell me what you would like to improve

A few lines is enough. I reply personally.