Sometimes the easiest way to understand what’s changed is to walk through a real type of scenario, step by step, and see how the output differs. This post does exactly that, taking a moderately complex SMSF scenario and showing how Charli’s agentic reasoning handles it differently from a straightforward policy lookup.

(Note: the scenario below is illustrative, built to demonstrate the reasoning process, not a real client file.)

The scenario

A client wants to purchase an investment property through their SMSF. They’re self-employed, with income verified via alt-doc, and the security property sits in a postcode that several lenders treat differently, some classify it as regional with reduced LVR caps, others treat it as standard metro fringe. The client is targeting an LVR that’s comfortable under a metro classification but tight under a regional one.

On the surface, this looks like four separate policy questions: SMSF lending criteria, alt-doc income requirements, postcode classification, and LVR caps. That’s exactly how brokers have historically had to approach it, four queries, four answers, then manual reconciliation.

Step one: identifying the factors

Charli’s reasoning process starts by breaking the scenario into its component parts, the same way an assessor would when first reviewing a file: SMSF structure, alt-doc income verification, security postcode, target LVR, and loan purpose. Nothing here is unusual on its own, it’s the combination that creates complexity.

Step two: checking each factor independently

Each factor is checked against current policy individually. Which lenders support SMSF lending for this property type? Which lenders accept alt-doc income for SMSF borrowers specifically (a narrower list than alt-doc generally)? How does each relevant lender classify this postcode? What’s the maximum LVR under each classification?

This step alone is where a lot of broker time has historically gone, not because any single answer is hard to find, but because finding all four reliably, and keeping them current as policy shifts, takes real effort.

Step three: cross-referencing

This is where the agentic model does its real work. Rather than presenting four separate answers, Charli checks how they interact. A lender with strong SMSF alt-doc policy might have a conservative postcode classification that drops the achievable LVR below what the client needs. A lender with a more generous postcode classification might have stricter SMSF documentation requirements that don’t suit this client’s structure. Charli surfaces these interactions directly, rather than leaving them for the broker to discover by manually comparing four separate answers.

Step four: comprehensive guidance

The output isn’t “here are four policy facts.” It’s a reasoned shortlist: which lenders genuinely fit once SMSF criteria, alt-doc treatment, postcode classification, and LVR are all considered together, with the specific friction points flagged for each option that’s close but not quite clean. That’s the kind of output you can take directly into a client conversation, or use to brief a BDM on exactly where you need clarification.

What this would have looked like before

Under the previous model, getting to the same place required running each of the four queries separately, then doing the cross-referencing, postcode against LVR, SMSF criteria against alt-doc treatment, manually. Accurate, but slower, and more dependent on the broker remembering to check every interaction rather than just the obvious ones.

What the broker still needs to do

None of this removes the broker from the equation. Charli’s reasoning gives you a well cross-referenced shortlist and the specific friction points to watch, but confirming the client’s actual fund structure meets SMSF trustee requirements, verifying documentation against the chosen lender’s exact checklist, and managing the client conversation about why one lender fits better than another remain squarely broker work. What’s changed is how much of the preliminary cross-referencing you need to do yourself before you get to that point. The four-query, manual-reconciliation version of this scenario could easily consume twenty or thirty minutes of careful checking; describing it once and reviewing Charli’s reasoned output is a different order of time investment, freeing up that time for the parts of the file that genuinely need your judgment.

Why case studies like this matter for building trust in the tool

It’s one thing to be told a tool reasons better than before. It’s another to see, step by step, what that reasoning process actually does with a scenario that has real complexity. That’s the value of walking through an example like this rather than just describing the capability in the abstract, it gives you a concrete sense of what to expect, and what kind of scenario description will get you the most useful output, before you try it on a live client file.

Why this case study matters beyond SMSF

SMSF scenarios are a useful illustration because they tend to stack several complicating factors at once, but the same reasoning process applies to any multi-factor scenario, complex income structures, near-threshold LVRs, postcode-sensitive deals, or combinations of all three. The pattern is consistent: identify, check, cross-reference, synthesise.

Try it yourself

If you’ve got an SMSF, alt-doc, or postcode-sensitive scenario on your desk right now, this is a good one to test the upgrade on directly. Describe the full picture to Charli in one go, the way you would explain it to a senior assessor, and see how the reasoning plays out. Start at cynario.ai/charli.

Author – Charli