Common questions.
Who or what produces the diagnosis?+
AI models score your work against a structured data-engineering rubric. The models are a trust mechanism, not the product – what matters is the rubric, the six skill areas, and the evidence linking. Where models disagree, we show the split and our confidence rather than averaging it into one number.
Is this a hiring prediction?+
No. It’s a diagnosis of your work across six skill areas – not a verdict on you and not a prediction of whether you’ll be hired. Levels describe your work, never you. We are not yet validated against hiring outcomes; calibration is in progress and we’ll show it as it grows.
How do I know the diagnosis isn’t just made up?+
Every finding is designed to cite evidence in your own submission – the exact lines, queries, or design choices it’s about – so you can audit the reasoning instead of trusting a black box. Undemonstrated skills are marked “not assessed,” never invented to fill a scorecard.
What if my submission didn’t exercise a skill?+
That skill area is marked “not assessed” – kept visibly separate from “attempted and lost points.” It’s a path forward, not a failure: the report shows how to demonstrate it next time.
What do I submit, and is it private?+
A curated question set you run in the browser, or your own repo. Your submission is private to you. Free to start – you can see where your work stands before deciding to go deeper.