What this looks like for a financial advice firm

This is an illustrative scenario, not a client engagement. It shows how the approach in the case study above would apply to a financial advice practice. Every figure here is an estimate offered for discussion, not a measured result.

The business

An Australian financial advice firm: a handful of advisers, a support team, and years of accumulated client files, statements of advice, compliance records, and product research.

The problem it would address

Advice firms carry the same shape of problem as a law practice — high-value knowledge, trapped in a file structure organised by client rather than by question.

• Preparing a statement of advice means recalling how the firm has handled a comparable client situation before, and finding the file it lives in.
• Compliance obligations mean the reasoning behind past advice has to be reconstructible, sometimes years later.
• Product and strategy research gets redone because nobody can find the analysis from eight months ago.
• Client information is exactly the kind of sensitive data that cannot be pasted into a public AI tool — so the staff who would benefit most are locked out of AI entirely.

What the approach would look like

The same two-layer design. A knowledge layer of de-identified firm expertise — strategies, product classes, regulatory concepts, advice frameworks — and a client layer holding the confidential files, linked into it.

The critical addition for a financial practice is access control, defined before anything is built: a support-team member's assistant should be able to search advice precedent without ever reaching a client's personal financial position. Governance is a design input, not a policy document written afterwards.

Delivery would run as a discovery engagement first — a week mapping where advisers actually lose time — before any system is built.

The kind of result to aim for

Realistic targets for a practice of this shape:

• Advice-precedent lookup reduced from a folder search to a query.
• Research not repeated because prior work is findable.
• Compliance reconstruction supported by a searchable record of what was advised and why.
• AI usable on real client work, inside a private environment, rather than banned outright.

The honest version of this: the value depends entirely on how the firm actually operates today, which is what a discovery engagement exists to find out.