What we do

Setting the direction. Data, AI and digital strategies, operating models, business cases, and the framing of reform and transformation initiatives.

Shaping the work. Forecasting demand, modelling populations, analysing scenarios and building the analytical evidence base before programs are designed.

Running the work. Digital services and AI in production, data engineering and pipelines, dashboards and reporting, process automation, and analytical prioritisation.

Measuring the work. Evaluation and outcomes measurement using advanced quantitative methods and linked data.

Developing a whole-of-state family violence data strategy

Ten-year cross-agency data strategy developed through consultation with more than 100 stakeholders across 10 agencies, guiding collection, performance monitoring, and reporting across the domestic, family, and sexual violence system.

Modernising the data backbone of a $2 billion primary care program

Automating data processes, developing operational dashboards, and supporting annual reporting for a major Commonwealth program spanning 31 regional primary care networks across Australia.

Risk indicators for disability scheme oversight

Development of risk indicators and market risk assessments to support scheme oversight and integrity.

“What is the strategic problem we are actually solving?”

We test the assumptions behind the brief with your senior stakeholders and reframe the problem, so the work targets the decision that matters rather than the one first described.

“What is the case for change, and what will it cost?”

Business cases and economic appraisal that size unmet need, cost the options, and set out funding scenarios a central agency will accept.

“How do we set up to deliver it, and know whether it worked?”

Operating model design that names roles and decision rights, paired with evaluation and outcomes measurement that tests whether the reform delivered.

“Where can AI genuinely improve outcomes?”

An AI strategy that separates the use cases with real value from the ones that only demonstrate well, sequenced against your data and capability.

“How do we adopt it without creating unacceptable risk?”

Responsible AI governance and assurance covering security, data sovereignty, human oversight, and the evidence you will need under scrutiny.

“What does a working solution actually look like here?”

AI solutions designed and implemented in your own environment and tested against real cases — generative systems such as document intelligence, agentic workflows, and machine learning models in production.

“How do we get something working in front of users, quickly and safely?”

A blended product team with a success measure agreed before the first sprint, something shipped inside the first cycle, and staged releases suited to government risk settings.

“Will it work with the systems we already have?”

Services designed around your existing platforms, data and security controls, and built in your own environment rather than bolted on beside it.

“How do we keep improving it once the project team leaves?”

Your staff pair with ours throughout, so your product manager can run the next cycle unassisted and Ember's share steps down as yours steps up.

“How much of this work still has to be done by hand?”

Process mapping and current-state analysis that quantifies the manual effort and shows where it accumulates.

“Where should we redesign rather than automate?”

Workflow redesign that removes steps first, so you are not automating a process that should not exist.

“What can we automate safely?”

Automated reporting and document generation, and AI-enabled task automation where the task is repetitive and the output is checkable.

“We hold a lot of data. Where should we invest first?”

A costed, sequenced roadmap that ranks initiatives by value and feasibility, so the first year of investment is defensible.

“Who is accountable for this data, and under what rules?”

Governance frameworks and operating models that assign stewardship, classify holdings, and set the guardrails for use.

“Why does every report start with manual data wrangling?”

Designed and automated pipelines that take raw administrative data to analysis-ready, on platforms sized for the load you expect and the security and sovereignty rules you operate under.

“Can we trust the numbers coming out of it?”

Linkage, matching and quality controls built into the pipeline, ensuring you get the same set of numbers each time.

“Can we find the signal in the data we already collect?”

Predictive and machine learning models built on your administrative data, tuned to the decision they inform.

“What can we learn by joining datasets we hold separately?”

Linked dataset analysis at state and national level to unlock cross-system insights, supported by rigorous quality and privacy handling assurances.

“Is the effect we are seeing real?”

Statistical testing and evaluation — including free text analysed through natural language processing — that establishes what the data can and cannot support.

“Can leadership see how the program is actually tracking?”

Executive dashboards built around the decisions your leaders make, showing the few measures that should change what they do next.

“How do we monitor delivery and meet our reporting obligations?”

Program monitoring and oversight reporting that serves internal management and statutory reporting from the same trusted source.

“Can people answer their own questions without asking us?”

Interactive, self-service analytics and public-facing reporting designed for people who are not analysts.

“How much of this service will be needed, and where?”

Demand forecasts at regional, state and national scale, built from population projections and observed utilisation, including the workforce required to meet them.

“What will this reform cost, and who will it reach?”

Cost projections and microsimulation that trace a policy change through to the people affected and the financial position over time.

“What if our assumptions are wrong?”

Scenario and sensitivity analysis showing which assumptions actually move the result.

“Where is the risk in a program this size?”

Risk detection models and provider or participant profiling that surface patterns that are difficult to identify manually.

“Which of these cases are worth acting on?”

Prioritisation and triage frameworks that rank cases by risk and likely return, so finite compliance effort goes to the right places.

“Can we defend how this decision was made?”

Automated decision tools built for proportionality and explainability.

We build relationships with clients and stakeholders, working together to achieve shared goals and successful outcomes.

We uphold excellence in all deliverables and services, maintaining high standards and continuous improvement.

We work embedded in client teams and build their data, AI and digital capabilities through pairing, training, mentorship, and resources.

We utilise a toolkit of consulting techniques to communicate results, create compelling presentations, and craft professional documents.