JevCompositeScore¶

Several score dimensions combined into one number with weights you control.

For ordering, never for gating

The argument this SDK makes for Jev over a single rubric prompt is that independent checks must be thresholded independently. An answer that is fluent and on topic but quotes an impossible temperature has to fail on plausibility alone, and any average hides exactly that.

So do not build a pass/fail decision on a composite. JevJudge thresholds each criterion separately and is the right tool for that.

What it is genuinely for¶

Ranking a field of candidates that have all already passed: which five of forty applicants to read first, which support ticket to work next.

There the dimensions are not pass conditions, they are preferences, and the weights are the policy. Keeping them in code means a disappointing ranking is fixed by changing a weight rather than by rewording a question.

Quick Start¶

JevCompositeScore engineering = JevCompositeScore.builder()
    .weight("python_depth",    0.40d)
    .weight("system_design",   0.40d)
    .weight("team_leadership", 0.10d)
    .weight("generalist",      0.10d)
    .build();

candidates.sort(Comparator.comparingDouble(c -> -engineering.of(c.response())));

Normalisation¶

Each dimension is divided by its rubric's own height (Score.maxLevel()) before weighting, so dimensions with different numbers of levels are comparable:

double normalised = JevCompositeScore.normalise(response.score("python_depth"));  // 0.0 – 1.0

With weights summing to 1, the composite is itself in [0, 1].

Same data, different profiles¶

The point of keeping weights in code is that one set of answers serves several rankings:

JevCompositeScore ic = JevCompositeScore.builder()
    .weight("python_depth", 0.40d).weight("system_design", 0.40d)
    .weight("team_leadership", 0.10d).weight("generalist", 0.10d).build();

JevCompositeScore manager = JevCompositeScore.builder()
    .weight("python_depth", 0.15d).weight("system_design", 0.20d)
    .weight("team_leadership", 0.40d).weight("generalist", 0.25d).build();

One call produces the answers; two weightings produce two shortlists.

Builder Configuration¶

Builder method	Type	Default	Description
`weight(String, double)`	—	—	A score criterion's share of the composite.
`allowUnnormalisedWeights(boolean)`	`boolean`	`false`	Permit weights that do not sum to 1, which puts the composite's range somewhere other than `[0, 1]`.

build() rejects weights that do not sum to 1 unless you opt out, because weights that do not sum to one are usually a typo rather than an intention.