Care Policy Concierge | Real Minds AI

Where it works well

It takes the whole "where is it written" lookup load off your policy and compliance people.

Best for a multi-site provider where staff can't walk to the quality-and-compliance officer's desk, and "what does our policy say" should always resolve to the current approved document.
It cites the document, version and clause it used (the demo lights up Records & Information Mgmt Policy v4, §6.3) so the staff member reads the actual policy, not a paraphrase.
The recaptured hours go back to your compliance lead and People & Culture team for the judgement work — interpreting an ambiguous case, drafting a new procedure, handling an exception.

The invisible cost in an aged-care provider is the lookup tax: the same handful of operational questions — records retention, the medication-error escalation, which form for a SIRS reportable incident — land dozens of times a week on whoever is seen as the policy authority, and the answer usually comes from memory, not the current document.

Where it works badly

It is only ever as right as the corpus you give it — and it cites a stale or contradicting version with full confidence.

It can't reason across documents or apply a threshold to a real case — "does this resident's fall meet the SIRS reportable threshold" is a judgement the policy text can't make for you.
If your policies live as old versions scattered across shared drives, it will surface a superseded medication procedure as readily as the live one.
A handbook that hasn't tracked the 1 November 2025 strengthened Standards will be quoted as if it had — the tool can't tell it's out of date.

The honest test

Pick five questions your staff actually ask and check whether each has a single, current, approved document that answers it in writing — if most don't, the document work comes first; the retrieval layer can't manufacture an answer that isn't written down.

Index three versions of the incident SOP, or two policies that quietly disagree, and it retrieves and cites one of them with the same 94%-match confidence it shows for a clean answer. It has no way to know which document is current or authoritative. That is the trap.

What it doesn't do — and shouldn't

It surfaces and cites what the approved document says. A person decides. That boundary is deliberate.

WHAT IT DOES

Names the document, version and clause it relied on, plus the verbatim quote

Shows a match signal and a "grounded in N sources" line

Refuses with zero sources and escalates to a person when nothing clears the threshold

WHAT IT WON’T

Decide what the policy should be, or write a new one

Certify that following the document satisfies the Aged Care Act

Apply a threshold to a real incident or interpret a grey case

Aged care is regulated by the Aged Care Quality and Safety Commission, which runs the Serious Incident Response Scheme — a confidently wrong answer about a reportable incident or medication management has real consequences for an older person and for the provider. So a Flag routes a wrong or stale answer to the document owner as a correction to the source, not a one-off, and the accountable person stays on anything consequential.

What your data has to look like

One current, approved version of each policy — clearly owned, with a visible version and date.

Typical readiness

across orgs we see, before the first job

One authoritative version of each policy

Needs shaping

A named owner for "what's current"

Needs shaping

Visible version and update date on every document

Usual weak point

Documents tracking the strengthened Standards

Usual weak point

Documents in a machine-readable form to index

Usually ready

The real first job

Most providers have the documents but not the discipline around them — three versions of the SOP, no single owner for "what's current". Establishing document control so there is exactly one live, owned version of each policy is usually a larger and more valuable piece of work than the retrieval layer on top — and it's about how you capture and approve documents, not a tool you buy. The concierge is only as trustworthy as the library beneath it.

Right fit if…

Multi-site provider whose staff can't walk to the compliance officer's desk

You have a genuine, approved document library staff need but don't read cover to cover

One current, owned version of each policy, with visible version and date

The same operational questions get asked dozens of times a week

Walk away if…

Your policies live as several versions scattered across shared drives

No single owner for "what's the current version"

Most questions need judgement across documents, not a single cited clause

You want a tool that certifies compliance with the Aged Care Act for you

Open questions

The worried-buyer questions, answered straight

It only answers from chunks that clear a relevance threshold, and it shows the document, version and clause it relied on plus the verbatim quote — so the staff member checks the source, not just the summary. When nothing in the corpus matches (the demo’s staff-cryptocurrency question) it refuses, returns zero sources, and escalates to a person rather than inventing a rule. In a sector run by the Aged Care Quality and Safety Commission under the Serious Incident Response Scheme, the safeguard is that the answer is never separated from its source, and a person stays on anything consequential.

Not well, and it will make the mess visible fast. The tool answers from whatever you index, so if two policies disagree it can cite the wrong one, or surface a superseded medication procedure with full confidence. Sorting out which document is current and authoritative — one version of each, clearly owned — is usually the real first job in an aged-care provider, and it’s the work we help with before the retrieval layer earns its keep.

No. It answers the high-volume “where is it written” questions so your compliance and People & Culture people aren’t the lookup desk, which recaptures their time for judgement work — interpreting an ambiguous case, drafting a new procedure, handling an exception. It cannot decide what the policy should be, only repeat what the approved document says, and it routes anything it can’t answer to a person — the demo escalates to People & Culture.

As current as your approved documents are — the tool is only ever as right as the version you’ve indexed. It shows the version and update date on every citation (the demo cites Records & Information Mgmt Policy v4, updated Feb 2026), so a stale answer is visible rather than hidden, but it has no way to know a policy predates the 1 November 2025 strengthened Standards. Re-indexing when a document is approved, and retiring superseded versions, has to be part of your document-control process.

The corpus is your approved documents only — the demo answers from those five documents and nothing else. A production build runs against a private retrieval index you control, not a public model’s training data, and staff questions are operational records you’d treat under your own Records & Information Management Policy. We scope data residency and access with you up front, because policy content and the questions staff ask about it are sensitive.

It answers strictly from your approved policy corpus and never beyond it: every answer carries the document, version and clause, and where nothing clears the threshold it refuses outright rather than blending a plausible answer from the open web. The refusal is the point — an intranet search returns the closest document; this returns the cited clause or nothing, and routes the gap to a person.

What it takes to build

3–4 weeks · 4 phases

Reused from template~70%

Bespoke to this skin~30%

stack · Claude · private RAG · review UI

What it would cost

Fixed scope, fixed price, fixed dates.

Bite-sized first piece

One contained change, low risk

Pilot build

Most builds land here

Embedded support

Scale on proof

Considering this for your org?

The honest place to start is a bite-sized first piece — one contained change, low risk. Tell us where the lookup tax hurts; we'll play it back, scope it, and show you what's possible against your own documents.

Book a call How we work