Jev versus risk rules: a practical comparison
Risk rules and scorecards encode credit, fraud, or ops policy as points and cutoffs. Jev does not become the scorecard. It fills the cells that used to be a junior analyst reading comments.
Unofficial sibling of the rules-engine page, focused on risk scorecards rather than generic RETE. docs.typesafe.ai.
Comparison scope
DTI, utilization, and country holds stay numeric. “Adverse media narrative” or “support chat implies first-party fraud” can be Jev inputs to the same scorecard.
Criteria that decide the architecture
| Axis | Jev (System One) | Risk rules |
|---|---|---|
| Cutoffs | Your code after Jev | Wins |
| Explanations to regulators | Question text you wrote | Rule ids |
| Unstructured adverse signal | Noul/Score | Keyword hits |
| Amount vs limit | Code | Code |
Decision quality and control
If an examiner asks “why declined?”, “Jev said so” is not an answer. “Rule R12 fired because the language Noul exceeded 0.8 on criterion X we documented” might be — talk to counsel.
Integration trade-offs
Scorecard rows: some computed, some Jev. Persist criteria version + model id + probabilities.
TypeSafe’s public models page lists jev-1.13 at $0.042 per million input tokens with output tokens free — a vendor claim, not a jev.pro measurement. Confirm on the models page before you bid.
When each approach fits
Prefer Jev when
- A scorecard row is a language judgment
Prefer Risk rules when
- The row is a number you already trust
Same pattern as rules engines, with stricter audit needs.
What this page does not claim
-
Not credit advice.
-
No PD/LGD claims.
-
Schema-safe output is not the same as factual correctness.
-
No independent bake-off numbers live here.
FAQ
Is Jev explainable AI? You can show criteria. That is not SHAP. See vs explainable models.
Can I put the whole bureau file in state? Minimize. Large irrelevant state hurts accuracy and bills tokens.
Disclaimer
This is an independent unofficial site and is not affiliated with TypeSafe AI; official documentation is available at https://docs.typesafe.ai. Never treat jev.pro as TypeSafe official documentation. We do not sell, issue, or proxy API keys.
Hub: Comparisons. Siblings: vs rules engines, vs fraud scores, vs compliance checklists. Canonical: https://docs.typesafe.ai.
Sources
Public TypeSafe or adjacent documentation only. No private claims.