About

Why SPARK exists.

Knowledge-work jobs are being reshaped by AI tools faster than the assessments traditionally used to evaluate fit have kept up. Multiple-choice tests measure recall. SPARK measures how a person reasons in writing through situations that have no single correct answer — the kind of work AI tools amplify rather than replace.

What SPARK measures

Six capacity dimensions that load on performance in AI-augmented work: Reasoning Under Ambiguity, Learning Velocity, Communication Architecture, Creative Problem Decomposition, Judgment Calibration, and Stakeholder Navigation. Each is anchored in published I-O psychology and cognitive science research; methodology lays out the construct definitions and the citation map.

Who built it

SPARK is a product of Defiant Solutions LLC, a Tustin, California consultancy focused on workforce capability for AI-augmented work. The scoring system, the rubrics, and the candidate-facing surface are built and operated by Defiant Solutions. The advanced AI algorithm that scores responses is paired with a human quality-review program. The I-O psychology partner work behind the rubric calibration is documented in the methodology page.

The state of the product

SPARK is currently in calibration with a limited cohort. The Reasoning Under Ambiguity rubric is fully calibrated against the v1.0.0 anchor responses. The five other dimensions are scored using v0.1.0 interim rubrics that will be re-calibrated against real respondent data through Phase 3 of the build. Until that work lands, every report card carries the rubric version that produced the score, and we explicitly flag the interim status on the page itself.

What we do not do

SPARK does not sell scores to employers. SPARK does not feed candidate data to third-party marketing or analytics. SPARK does not represent itself as a hiring credential, a performance management tool, or a compliance product. The intent is developmental feedback for the candidate, full stop.