We use AI. It is never allowed to decide anything about your money.
Plenty of people in finance have quietly started running your numbers through a chatbot. We built ours the other way round. The rules are written in code that cannot change its mind, and the software’s job is to record what a human did, then tell you what is still missing.
Finalist, Australian AI Awards 2026
Healthy Wealthy Investor was named a finalist in the Australian AI Awards 2026, AI Innovator for Property Investment, announced 7 September 2026 by Momentum Media.
Juan Jeffery · Strategic Property & SMSF Credit Advisor · Australian Credit Representative CR 464548 · Finsure ACL 384704
The three things we actually built
No vague talk about being powered by AI. These are the systems, by name, and what each one is for.
SMSF Concierge, the one you can use right now
Ask it a question about self-managed super funds, borrowing or property structure and it answers in plain English. Its answers are grounded in the actual superannuation and credit legislation alongside our own published education, and it shows you what it drew on so you can check it. The moment a question turns personal, it stops and tells you to speak to a human. That refusal is deliberate. It is the feature, not a limitation we are apologising for.
ProofMemo, the system that checks the file
When a lending file is put together, somebody has to confirm that every document is there, that the income figures agree with each other, and that nothing has been skipped. Traditionally that check lives in the broker’s head and leaves when the broker does. ProofMemo reads the documents, reconciles the numbers across them, and produces a dated record of what was checked. Most importantly it lists what it could not establish. On a recent file it found four identity forms with no applicant name and an unsigned declaration. To anyone flicking through the folder, that file looked complete.
The workflow engines, the unglamorous ones
Client onboarding and document reading, a serviceability calculator that uses the Household Expenditure Measure and the regulator’s buffers, the twenty step compliance process written into code, and an SMSF borrowing qualifier. Built between February and June 2026. None of them guess. Given the same documents they produce the same answer every time, which is the whole point.
Four numbers, measured not claimed
4,705
Automated tests that must pass before any change goes live. Measured 9 September 2026.
19
Steps in the assessment line, each one labelled with who acts there and whether it touches your rights.
0
Client documents that leave the practitioner’s own machine while they are being read and assessed.
Mar 2026
When the first gate blocking identity documents and tax file numbers from reaching any model went into production code.
Australia has six rules for using AI responsibly. Here is how we go on each one.
In October 2025 the National AI Centre published its Guidance for AI Adoption, built around six essential practices. There is no exam and nobody hands out a certificate. So the only honest thing to do is go through them one at a time and show our working, including the one we fail.
1. Decide who is accountable
Every step in the assessment line is labelled with who acts there: the broker, the software, or neither. The record says on its face that the professional judgement belongs to the human. It specifically refuses to certify that the best interests duty was met, because no piece of software can carry that duty for a person.
2. Understand the impacts and plan accordingly.
Our AI System Impact Assessment sets out who the system can affect, how, and the mitigation in place for each impact. It is published in full at /ai-impact-assessment/ and reviewed at every major release.
3. Measure and manage the risks
We keep a ranked list of the ways our own system could be beaten, sorted by how exposed it is, how much damage it would do, and how easy it would be. The list opens on our worst weakness rather than our best feature, and it doubles as the build queue. Fixing the top item is always the next job.
4. Share what matters
We publish what our system still cannot catch, and we publish the test results including the cases it fails. A safeguard nobody is allowed to inspect is just a claim.
5. Test it and keep watching it
4,705 automated tests run before any change goes live. One of them exists purely to fail the build if any step in the assessment line is caught making a decision on its own. Documents are sorted and cleaned before a model ever sees them, so identity documents, tax file numbers and account numbers are stripped out first.
6. Keep a human in control
The headline behaviour of our software is that it declines to act. It stops at the line between general information and personal advice and records that it stopped. If it meets a document laid out in a way it does not recognise, it halts the run rather than having a guess. A field it cannot read is marked unreadable and pointed to on the page, so a person can go and look.
Where we stand against the recognised standards
A first-party self-assessment against the international standards the field is measured by. We name what we conform to today and what we are working toward, and we hold no third-party certification yet.
ISO/IEC 42001, AI management
We have self-assessed against the international standard for managing an AI system. We conform today on how the system runs and how its risks are managed and monitored: the decision path has no model in it, the controls are tested on every build, and we keep a live risk register. We are working toward certification on the surrounding management system, and say so. Our AI System Impact Assessment is published in full.
ISO/IEC 27001, information security
The technical security controls are strong because the architecture is built for it: documents stay on the practitioner’s machine, access to raw identity data is gated, and the record is cryptographically signed against a key we publish outside the file. The full statement of applicability and the organisational controls are the work in progress.
SOC 2, an auditor’s report, not a claim
SOC 2 is an independent auditor’s report produced over a period of observation, not something an organisation declares of itself. We make no SOC 2 claim. We are working toward a SOC 2 Type 2 report over our hosted verification service, and will name it here when it is issued.
Where the line is, in writing
General information, never personal advice
Nothing produced by any system on this page is personal credit, financial, tax or legal advice. None of it knows your goals, your situation or your needs. Whether something suits your own fund is a question for your licensed adviser and your accountant.
The record belongs to the broker, not the software
ProofMemo records what a credit practitioner did. It does not make the professional call and it does not endorse one. Every record it produces says exactly that on its face.
We assist the human to carry out their duty, we do not take it from them
In real time, the software helps the licensed human carry out their duty to the best standard the market offers, and in compliance with the regulations, acts and standards that govern the work, every required check run the same way each time. What it never does is take that duty over, because a professional duty cannot be delegated to software and should not be. So the system runs locally, in the broker’s own browser, holding no second copy: the client’s data stays in the broker’s hands and the judgement stays with the person the law trusts to make it. We take full responsibility for what our software does and build it to strengthen the human’s hand, never to replace it and never to quietly shift the onus onto you.
We hold no AI certification, and we are not going to imply we do
Everything on this page is our own assessment of our own work, made in the form of a supplier’s declaration of conformity under ISO/IEC 17050-1. It is a first-party statement. It is not a certification and it has not been audited by anyone else. Work toward independent certification has started, and we will name it here on the day it is issued and not before.
Start with a question. Move to a conversation when you are ready.
Ask the SMSF Concierge anything you like. It costs nothing, it cites what it used, and it will tell you honestly when your question has moved past what it is allowed to answer.
When the question becomes about your structure, that is a conversation with a human. Book a strategy session at AeFin.as.me.
Healthy Wealthy Investor is a brand of Aubelia Enterprise Pty Ltd trading as AeFin, an Australian Credit Representative (CR 464548) of Finsure Finance and Investment Pty Ltd (ACL 384704). The material on this page is general information about how we build and govern software. It is not personal credit, financial, tax or legal advice, and it does not take account of your objectives, financial situation or needs. Page last reviewed 11 September 2026.