Who we are and how tibbi is governed

This product asks you to trust that the parts of it which could hurt someone have been checked by a person with the training to judge that. Here is who that person is, what she has the power to stop, and what happens when she is not available.

The organisation behind the service

Operating entity	Test It Before Bet It Ltd, a not-for-profit company limited by guarantee, registered in the Australian Capital Territory. tibbi is its registered business name.
ABN	42 683 104 927
ACN	683 104 927
Registered office	Canberra, ACT. Correspondence to PO Box 4127, Weston Creek ACT 2611.
Public liability insurance	$20 million, held continuously while the service operates. Evidence is provided to the funder on request.
Professional indemnity	$5 million, covering the content review work described below.
Board	Three directors, one of whom is independent of the delivery team and chairs the conflict of interest register.
Principal funder	ACT Gambling and Racing Commission, through the Gambling Harm Prevention and Mitigation Fund. See how tibbi is funded.

Why an ABN is on a public page

A service that collects a record of where people are financially vulnerable should be traceable to a legal entity that can be held to account, not to an app store listing. If you want to check who we are before you create an account, everything you need to look us up is above.

Details of organisations funded by the Commission are published on the Commission's own website, so this page is not the only place you can verify it.

The people

A small team. We would rather say that plainly than imply a department that does not exist.

🧭

Founder and product lead

Name to be added

A postgraduate background in information technology, experience building games and software, and a working knowledge of the behavioural psychology that betting products are designed around. Lost control of their own betting as a student and needed family help to clear it. That experience is why the simulation is uncomfortable rather than fun.

Decides: product scope and priorities. Cannot decide: whether sensitive content ships.

🤝

Lived experience and community liaison

Name to be added

Ran a hospitality business and lost money to betting and poker through it. Reviews the simulation against how the pull actually feels, and tests whether the warnings read as honest or as a lecture.

Decides: whether a mechanic is truthful to the experience. Cannot decide: clinical safety, odds, tokens.

🩺

Clinical adviser

Name to be added · content review · right of veto

Registered psychologist, counselling practice at Belconnen Hospital. Engaged by the hour to review the material this product puts in front of people who are already under strain. Her authority is set out in full in the next section, because a vague claim that "a clinician was consulted" is worth nothing.

Decides: whether sensitive content ships. Cannot decide: anything about an individual user.

📚

Sam Okafor

Content administration

Writes and maintains the lesson videos, the quiz bank and the facts shown in risk warnings. Sends sensitive material for review. Sees no user data.

📊

Priya Raman

Impact and measurement

Builds the pilot reporting. Works only with aggregated figures, and only publishes a group's numbers when there are at least ten people in it.

🔧

Marcus Hale

System administration

Keeps the service running and holds the only role that can see an account and its cooldown status. Every use of that access is written to a log that he cannot edit, and changes that matter need a second approver.

What the clinical adviser signs off, and what she has nothing to do with

This is the load-bearing claim of the whole product, so it is written as a boundary rather than as a reassurance.

Cannot ship without her written approval

The facts inside risk warnings. Every statement shown to someone about to sell a simulated asset, with its source.
The critical-state screen and the screen where a run ends. Both reach a person at their lowest point in the simulation, and both must carry the support panel.
The AI companion system prompt and its crisis-recognition routing. This one is a release blocker, not a recommendation: without her sign-off the companion does not go out, whatever the deadline says.
Anything sent to a person during a cooldown. A badly worded reminder to someone who has just lost a simulated life can do real harm.

Approval is recorded against the specific version of the content. Change a word and it goes back into the queue.

Outside her role, by design

She never sees player data. Not names, not accounts, not one person's betting or stress readings. She reviews content, and she does not practise on our users. That boundary protects you and it protects her registration.
She does not write what she approves. Authoring and approval are separate people. Where the team is too small to hold that apart, the exception is recorded in writing rather than quietly ignored.
She has no access to odds, tokens or cooldown. None of the three can be changed by a clinical judgement, or by anyone else's.
She does not diagnose or treat anybody through this product, and neither does the product.

When she is not available

One reviewer is a single point of failure, and pretending otherwise would be the first dishonest thing on this page. Two arrangements cover it. A second registered psychologist is retained on the same terms and can approve in her place, and the register records which of the two approved each item. If neither is available, the queue simply waits: the release goes out without the unreviewed content rather than with it. There is no provisional approval, no "ship it and review afterwards", and no override held by the founder.

How the safety claims are verified before each release

The public statements on the home page — no deposit path, tokens cannot be bought, odds are never adjusted to a person, no bookmaker link, no timer in a quiz, no way to shorten a cooldown, Family cannot be restored — are technical constraints, not promises. Each one has a check that is run before a release goes out.

The full set of checks is run against the build, not against a description of the build.
The result of every check is written to a ledger with the date, the version and the person who ran it.
A failed check stops the release. There is no sign-off that waves one through.
The ledger goes to the funder with each progress report, and the checklist itself is published for other services to reuse.

The checklist is available here, in full, without asking us first.

Who can overrule whom

The clinical adviser can stop content. Nobody can overrule that, including the founder and the board.
The board can stop the project. It cannot approve content in her place.
A sponsor can stop its own campaign. It has no say over any other part of the product, and no say over what a lesson teaches.
Nobody at all can lift a cooldown, restore a lost Family asset, or issue tokens outside the two learning paths.

Conflicts of interest

The same organisation runs a harm prevention service and sells sponsored placements inside it. That is a real conflict and it is managed rather than denied.

Interests are declared to the independent chair and kept in a register reviewed each quarter.
Anyone with an interest in a sponsor steps out of the decision about that sponsor.
No sponsor money is spent on the learning, betting, cooldown or support layers, and no grant money is spent on sponsorship. The separation is set out here.

What this organisation is not

We are not a treatment provider, not a counselling service, and not a replacement for the ACT Gambling Support Service or Gambling Help Online. We route people to them by design, at the four moments in the product where it matters most. If you want the detail of that boundary written for a professional audience, it is on the page for ACT services.

Something here does not add up

Then tell us, and tell the funder if our answer is not good enough. Both routes are published, and using either one has no effect on your access to the service.