PMSI or GSA?

PMSI or GSA?
Let Jev tell you!

Input your financing statement wording (add a maximum of 3 collateral entries). Jev will tell you whether it's a PMSI, GSA, both, or Other, including its confidence & ambiguity score.

01 / INPUT

Collateral entries

1 of 3
Entry 01

Descriptions stay server-side and are sent to Jev only when configured.

FAQ

What can be classified as a GSA?

GSA requires the selected collateral type All present and after acquired personal property with an unrestricted grant. All present and after acquired personal property except... cannot be GSA: it may be PMSI when the rest of the wording limits the collateral to supplier-provided property, or Other when it does not. Broad wording entered under Intangibles, Negotiable instruments, or another collateral type also cannot produce a GSA.

What points to a PMSI?

Supplier, purchase-money, lease, hire, bailment, retention-of-title, or similar wording points to PMSI. A separate ALLPAAP entry limited only to proceeds of those goods remains PMSI. The same applies to wording such as “ALLPAAP except goods not supplied by us”, because that exception limits the grant to supplier-provided property.

When can it be both PMSI and GSA?

A financing statement can contain both. The combined result applies when one entry covers supplier-linked or purchase-money collateral and a separate entry creates an unrestricted ALLPAAP grant. An ALLPAAP entry limited only to proceeds of the supplier-linked collateral does not create a separate GSA, so that remains PMSI only.

What does the ambiguity score measure?

It is a caution score out of 100, not the percentage chance that the result is wrong. The tool uses the higher of two Jev signals: whether the wording appears conflicting or incomplete, and how strongly Jev favoured one classification over the others. A conflict between broad ALLPAAP wording and a non-ALLPAAP selected type sets a minimum score of 50. The explanation under each result identifies which factor drove that result.

Who or what is Jev?

Jev is TypeSafe AI's decision model. A regular large language model generates open-ended text one token at a time. Jev instead answers questions whose possible outputs are defined in advance. Here, it chooses PMSI, GSA, PMSI & GSA, or Other and separately scores ambiguity; this site applies its hard collateral-type rules and formats the result.

Why is Jev such a big thing?

It is designed to make small, structured decisions much faster and more cheaply than a general-purpose language model. TypeSafe reports typical Jev response times of 70–500 ms and a price of US$0.042 per million input tokens, with no output-token charge. Its published comparison puts frontier LLM responses in seconds and at higher token prices. Those are TypeSafe's figures, not a guarantee for every request; this site shows the actual latency and estimated query cost for each result.

Can Jev get it wrong?

Yes. Jev always returns an answer in the required format, but that does not make the classification legally correct. A choice weight such as 97% means Jev strongly preferred that label over the other labels it was offered; it is not a 97% guarantee that a court or practitioner would agree. This site applies hard rules after Jev responds and uses the ambiguity score to expose conflicts or uncertainty. Treat the result as triage and check the security agreement and registration wording before relying on it.

What happens if Jev is unavailable?

The tool uses deterministic local rules covering broad ALLPAAP grants, supplier restrictions, proceeds-only ALLPAAP entries, and type-description conflicts. The result card shows which engine produced the indication.