ZazenCodes demonstrates how to query TypeSafe's Jev model using the Python SDK. The tutorial covers Jev's core decision primitives—Noul, Choice, and Score—along with parallel multi-question evaluation and an automated support ticket triage loop.
Original by ZazenCodesGetting startedIntermediate20 min 58 secPublished
Jev is a non-generative, System 1 decision model that enforces schema adherence across core primitives like Noul, categorical choices, and ordinal scores.
Multiple questions of differing types can be evaluated against input state in a single parallel call, returning probability distributions.
Autonomous triage workflows can use Jev's output confidence scores to trigger deterministic actions or escalate low-confidence cases to reasoning agents.
Worth knowing
The triage loop uses mock actions rather than live production integrations, and Jev lacks built-in web search, meaning out-of-the-box knowledge is constrained by its training cutoffs.