PolicySim

For independent advisors and RIAs

Annuity analysis that shows its work.

PolicySim projects a household year by year against real carrier contract terms, and records what it considered, what it assumed, and who signed off. Not a black box with a number at the end.

Deterministic, not indicative

The same household and the same assumptions produce the same projection, every time. Runs record the engine version and the rate-sheet snapshots that fed them.

Carrier terms, with provenance

Rates come from carrier documents, and every figure carries the page and box it was read from. Where a document is unclear, the gap is reported as a gap.

A human approves, or nothing ships

A ranked comparison is a draft until a named, licensed reviewer approves it. The record keeps both the decision and the checks that ran.

How it works

Five stages, and you can see into all of them.

Carrier documents become typed facts before they become a catalog, and the catalog is what the engine reads. Nothing is hand-keyed, and nothing is a number somebody remembered.

1 · HarvestCarrier rate sheets and contract documents, collected on a schedule.
2 · ExtractDocuments parsed to typed facts — caps, rates, riders, surrender schedules.
3 · CatalogFacts projected into products, dated and versioned, with state overrides.
4 · ProjectHousehold modelled year by year against those terms and your assumptions.
5 · ReviewComparisons ranked, checks run, and a licensed human decides.

Scenarios you can move

Change a premium, a retirement date, or an inflation assumption and watch the curve respond immediately. Exploration records nothing — saving a run is a separate, deliberate act.

Your data stays yours

Every account is isolated at the database level rather than by a filter in application code, so a query that forgets the scope returns nothing rather than somebody else's client.

Governance

What this software will not do.

PolicySim is built against a written best-interest protocol, and the parts of it that constrain us are the point. Some of these are things competitors advertise.

It will not tell you a product is "best"

A ranking is only meaningful against a stated comparison scope, and the output always names that scope. If one product was tested, the output says no meaningful comparison was performed.

It will not present an illustration as a recommendation

Illustrating a product you chose and ranking a shelf are different questions, so they are different screens with different records. Neither is quietly relabelled as the other.

It will not score on your compensation

Compensation is not modelled in the catalog and carries no weight in scoring. A check asserts that rather than assuming it, so the day it stops being true, something says so.

It will not hide a tradeoff it found

Material downsides travel with the result. A check that could not run reports that it did not run — never a green tick for diligence nobody did.

PolicySim is analysis software, not an advisor. It does not provide investment, tax, or insurance advice, and nothing it produces is a solicitation or a recommendation to any consumer. Suitability and best-interest determinations remain the responsibility of the licensed professional using it — which is why the software requires one by name before a comparison can be released.

Pricing

Per seat, monthly. Cancel whenever.

One account per firm. Every person in it gets their own login; an administrator decides who may approve advice.

Pricing shown at checkout
  • Household modelling and the projection sandbox
  • Full carrier catalog with document-level provenance
  • Comparisons, decision records, and the reviewer queue
  • Unlimited projections — nothing is metered per run

You will set your password after checkout — we email you a link. Payment is handled by Stripe; no card details reach us.