Small decisions. Interesting possibilities.Submit contentSubmit
jev-recruiter preview

jev-recruiter

A Jev powered LinkedIn recruiting agent. Watch it browse relevant profiles, save links, and review evidence against your hiring brief.

Added to Jevfast

How it uses Jev

The recruiting loop uses Jev for observed browser operations, professional-title screening, and profile assessment. For each criterion, Jev chooses a status and an indexed observed excerpt; code resolves the quote from captured text and validates it before saving the assessment. Morph is an explicitly configured alternative provider.

What Jev decides

Candidate title checkExample answers · not a recorded Jev response · Source ↗
Question 1 · p1 status
YOUR APP
INSTRUCTION

Does the first candidate’s own professional title merit opening the profile?

STATE

The requested role and bounded, observed title excerpts from a candidate card are supplied as state; this is a navigation relevance check, not a hiring decision.

JEV · CHOICE
  1. relevant
  2. unrelated
  3. unknown

App workflow

  1. Set a brief

    The user edits a LinkedIn search, required/optional criteria, limits, and a target number of potential matches.

  2. Navigate and screen titles

    Jev chooses observed browser operations; observed profile cards are screened for an explicit relevant professional title before opening.

  3. Assess profile evidence

    After a profile is opened, Jev selects a criterion status and indexed excerpt; code checks that the quote occurs in the observed profile text.

  4. Review and export

    The user reviews saved profiles and evidence, shortlists or passes candidates, and can export the run as JSON.

Why it is interesting

It preserves source excerpts beside each criterion decision and downgrades unsupported or invalid quotes to unknown. Navigation choices are also constrained to observed targets and checked for page freshness before execution.