Deterministic, replayable calculations
RiGEL's engine is deterministic by design: the same inputs and rule version always produce the same outputs, making every scenario traceable, testable, and defensible.
Deterministic vs probabilistic
- No Monte Carlo simulations or opaque risk scores.
- Each scenario follows a defined sequence of rule evaluations.
- Inputs, assumptions, and rules are visible and reviewable at every step.
Replay and independent verification
- Advisors and reviewers can rerun the same scenario with the same inputs and rule set and receive identical outputs.
- Any change to assumptions, dates, or jurisdictions results in a clearly documented new run.
- This supports internal QA, peer review, and regulator questions such as "How did you get that number?"
Controls and guardrails
- Input validation to catch missing or inconsistent data before calculations run.
- Rules for edge cases (e.g., partial years, multi-jurisdiction estates, treaty tie-breakers).
- Versioned rules so firms can align engine behaviour with their internal compliance sign-off.
For engineers & compliance teams
How the deterministic engine behaves
At a high level, RiGEL executes estate scenarios as a series of deterministic rule evaluations rather than as free-form scripts.
- •Rules are grouped into jurisdictional modules (e.g., CRA, IRS, provincial probate) and executed in a consistent order.
- •Each step records inputs, outputs, and applied rule IDs for traceability.
- •Scenario runs can be versioned against specific rule sets, which supports change-control and internal model governance.