Small decisions. Interesting possibilities.Submit contentSubmit

How it uses Jev

The PostgreSQL extension accepts a text or JSON state plus typed question definitions, sends a System One request, and returns Jev’s typed answers to SQL. Convenience functions expose Choice classification, Noul detection, and Score evaluation.

What Jev decides

Illustrative support-ticket classificationExample answers · not a recorded Jev response · Source ↗
Question 1 · label
YOUR APP
INSTRUCTION

Which team should handle this?

STATE

A SQL example supplies a support-ticket text as state and asks which team should handle it. The criteria are caller-provided.

JEV · CHOICE
  1. billing
  2. technical
  3. sales

App workflow

  1. Prepare SQL input

    Supply state and question instructions with criteria or a rubric through an extension function.

  2. Build and send request

    The extension serializes state and typed questions for the configured Jev model.

  3. Read typed answers

    Consume the returned Choice, Noul, or Score answer in SQL.

Why it is interesting

It brings Jev’s bounded judgments into database workflows as SQL calls, so applications can classify, detect, or score values without moving that logic into a separate app service.