Skip to main content

Documentation Index

Fetch the complete documentation index at: https://help.rytz.com.au/llms.txt

Use this file to discover all available pages before exploring further.

The Settlement Planner has an AI analytical layer that runs across all four steps of the framework. It reasons about your inputs, surfaces patterns, suggests reasoning, and produces structured second opinions on contested questions. Click AI suggestions at any point in the Settlement Planner to access it.

​
What it does

Five distinct functions across the framework:

Pool completeness suggestions

Reads your Asset Pool and suggests what might be missing. “You have a family home; is there a mortgage? You have super for one party; what about the other?”

Contributions reasoning

Given the matter shape, suggests how the contributions analysis typically plays out. References analogous published cases and explains the case-law line.

Future-needs framing

For each s75(2) factor, suggests typical magnitudes from analogous cases. Surfaces factors you may have under-weighted or overlooked.

Just-and-equitable second opinion

Reads your proposed Step 1–4 result and assesses whether a senior solicitor would call it Reasonably fair, Aggressive, Conservative, or Unduly harsh. With reasoning.

Negotiation framing

For offers logged in My Offers, suggests how each compares to the analysed position and what a counter-strategy might look like.

​
How suggestions are generated

The AI suggestions layer runs Claude Opus over:
  • Your full Settlement Planner state across all four tabs
  • Your Master Case File (matter shape, parties, children, FV disclosure, ATSI status)
  • Your Evidence Portfolio (financial documents, supporting evidence)
  • The post-10-June-2025 Family Law Act framework
  • The case-law corpus from the Legal Research Library (2,389+ Gold Standard vectors)
  • Anonymised analogous-case patterns
Each suggestion is grounded in a specific authority or pattern. Hover or click “Why?” on any suggestion for the reasoning.

​
When to use suggestions

Three specific moments:

At the end of each step

After completing your input on a step, run AI suggestions for that step before moving to the next. Catches gaps and over/under-weightings before they cascade.

Before sending an offer

Run suggestions across the full framework before any offer goes to the other party. Final sanity check on whether the position you’re proposing aligns with the analysis.

Before lawyer consultation

A solicitor’s hour is more productive if you’ve pre-tested your analysis against AI suggestions. Suggestions catch the basic issues; the solicitor focuses on subtle ones.

When stuck

For decisions you’re turning over (whether to accept their offer, whether to escalate to court, whether to rebalance super), suggestions provide structured reasoning that breaks deadlock.

​
Reading suggestions well

Several patterns:
Treat suggestions as a structured second opinion, not as advice. They are calibrated to surface the kind of reasoning a senior family-law solicitor would offer. They are not legal advice; the decisions remain yours.
Pay attention to “Why?” reasoning. A suggestion without grounding is less useful than one with. The grounded suggestions reference specific cases, sections, or analogous-matter patterns; treat the reasoning more seriously than the headline.
When suggestions disagree with you, ask why. The most-useful suggestions are the ones that flag a different view from yours. Read the reasoning; if it’s unconvincing, override; if it’s compelling, adjust.

​
Common suggestion patterns

Three typical surface patterns:
PatternWhat it usually means
Many suggestions on Asset Pool, few on ContributionsPool isn’t yet complete; address before moving on
Suggestions on Step 3 surface a factor you’d overlookedCommon; the s75(2) list is long and self-rep parties often miss factors that are present
Just-and-equitable verdict “Aggressive”The proposed split would face headwind in court; consider whether your analysis supports it or whether you’ve over-weighted contributions
Just-and-equitable verdict “Unduly harsh”The proposed split would likely be corrected by the court; re-examine

​
Where suggestions fall short

Honest limits:
  • Suggestions don’t know what you haven’t told the platform. A factual nuance not captured in the Master Case File or Evidence Portfolio is invisible to the AI.
  • Suggestions don’t predict outcomes. Family-law settlement outcomes are heavily fact-driven. The AI provides structured reasoning, not predictions.
  • Suggestions assume routine matter shapes. Highly unusual matter shapes (international elements, contested mental-health questions, complex business structures) are at the edges of the AI’s training; specialist input typically warranted.
  • Suggestions are not advice. As with the rest of the platform’s AI surfaces, this is information, not advice. See AI limits and safety for the boundary.

​
What’s next

Settlement Planner overview

Step back to the framework.

Step 4 — Just and equitable

The Fairness Analyser is part of the AI suggestions layer.

Offers and negotiation

AI suggestions on negotiation context.

AI assistant overview

The general-purpose AI surface — useful for follow-up questions on Settlement Planner suggestions.