One live view for a leading NDIS provider.

The challenge.

NDIS price limits are set nationally. A provider cannot price its way to margin, and the great majority of every dollar it earns goes straight back out as award-based wages for the support workers and nurses delivering care. That leaves exactly one lever: taking cost and risk out of administration. It is also why legacy providers keep exiting complex NDIS work - they cannot run lean enough to make the numbers add up.

This provider had roughly doubled revenue every year since launching, and was targeting $30M in the next financial year. But the operating model underneath had not kept pace. Finance, rostering, CRM, payroll and learning data all sat in separate tools, stitched together by manual exports and patch-job transfers. Nobody could see cash flow, onboarding, compliance and capacity in one place, so leadership answered questions by asking people rather than by looking.

The other constraint was mobilisation. Accepting a complex placement means standing up a furnished home, a vetted and matched team, and a full compliance pack, quickly. Every new participant and every new staff member is a fresh risk to be vetted, trained and matched. Done manually that is demanding at $20M. At $30M it does not work, and the only conventional fix - more coordinators and more managers - is the one thing the margin cannot fund.

What RUBIX did.

RUBIX started with the layer most AI projects skip. Bolting agents and chatbots onto disconnected systems produces demos, not outcomes, so the first eight weeks went into the foundation that makes automation reliable in a regulated industry. Working alongside the provider's operations, finance and compliance leads, we:

• Stood up a secure, sovereign Australian environment and connected the core source systems - finance, rostering, CRM, payroll and learning - mapping every data flow and eliminating the manual patch-job transfers between them.
• Built a governed data foundation joining those six systems into one source of truth, with row- and column-level security so a support worker, a coordinator and a director each see exactly what they should and nothing more.
• Delivered a live, role-based dashboard covering cash flow, onboarding, compliance and capacity in one view - queryable in plain English, so leaders get an answer without waiting on a report.
• Installed an enterprise-grade governance layer with security, compliance and AI explainability built in from day one, ahead of Australia's incoming AI obligations, so AI can be relied on where care and funding are at stake.
• Produced an automation roadmap with hard numbers: which manual processes carry the most cost and risk, which AI should handle first, and the projected return on each - ranked by ROI, not by what is interesting to build.

Delivery was fixed-price and fixed-scope, with the deliverables owned by the provider regardless of what came next. No open-ended engagement, no 18-month timeline, and no in-house IT department required to run it.

The results.

Leadership now opens one live view instead of assembling a picture from six systems and a series of phone calls. The data foundation underneath is governed, secured by role and sovereign, which means the automation work that follows starts from something trustworthy rather than from spreadsheets.

• Six disconnected systems replaced by one platform, with manual transfers between them eliminated.
• A working dashboard, live in under eight weeks - not a mockup and not a strategy document - showing cash flow, onboarding, compliance and capacity by role.
• An automation roadmap quantifying ~$645K in conservative annual impact: administration and coordination capacity redeployed to care, faster placement mobilisation on the state-funded side, fewer documentation and medication errors, and leadership time returned to growth.
• Governance and AI explainability in place before the automation build, so every AI-assisted decision is traceable to source.
• A foundation that absorbs the move from $20M toward $30M as volume, not as extra admin layers.

Why it matters.

Care providers are in an unusual position: demand is growing, pricing is capped, and the compliance burden is rising at the same time. Growth cannot be bought with headcount, because headcount is the cost the price limit will not carry. The only durable answer is infrastructure - a data foundation that gets the operating knowledge out of people's heads, an executive dashboard that makes the business visible in real time, and agentic AI that absorbs the repetitive admin once the data underneath it can be trusted.

The sequence matters more than the tooling. Get the data and the governance right first and everything built on top runs reliably, at scale, and stands up to a regulator from day one. Skip it, and you get pilots that demo well and deliver nothing.

A delivered RUBIX engagement, published with the client's identity withheld at their request. The annual impact figure is the Phase 1 roadmap's conservative projection, not a realised saving.