Abstract
A system is proposed herein for qualifying evaluators used to review generated technical material in versioned engineering artificial intelligence (AI) systems. The system creates controlled degradations from protected technical obligations, measures the technical dimensions for which each evaluator demonstrates competence, and limits promotion authority to current qualified dimensions. Dimension-specific authority rules prevent lower-authority votes from overriding controlling technical sources, while named dependency envelopes invalidate affected qualifications when relevant sources, platforms, releases, or evaluator configurations change. Unaffected evaluator capabilities and still-valid prior decisions remain usable, allowing localized promotion, rejection, owner review, or insufficient-evidence outcomes.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
C, Bharath Raj; Anandan, Mukund; and Gourav, Gourav, "MUTATION-QUALIFIED, AUTHORITY-SCOPED EVALUATOR ROUTING FOR VERSIONED ENGINEERING AI SYSTEMS", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/12352