Adaptive Mode

Adaptive mode gives the AI full competency anchors and lets it generate questions in real-time based on the candidate's responses. This is the platform's key differentiator.

How It Works

1. Employer selects competencies and weights (no question spine needed)
2. At interview start, the AI receives full 5-level anchors for each competency
3. The AI generates questions targeting level 3-5 evidence per competency
4. Based on responses, the AI probes deeper or moves to another competency
5. The AI self-manages coverage across all competencies
6. The scorer works identically — scores against the same anchors

Configurable Fields

All configured via admin at /admin/adaptive-prompt:

Question Generation

• Opening question strategy — which competency to start with
• Question formulation rules — constraints on valid questions
• Anchor targeting instructions — how to use 1-5 levels when generating questions

Probing

• Probing triggers — signals that trigger deeper probing
• Max probes per competency — hard cap (default 3)
• Probe types — clarifying, depth, ownership, reflection
• Escalation rules — when to probe harder vs. move on

Coverage

• Minimum coverage rule — must touch every assigned competency
• Coverage/depth tradeoff — how to balance breadth vs. depth
• Budget allocation — question allocation strategy (proportional to weights)
• Max competency share — max % of questions on any single competency (default 40%)
• Coverage tracking — self-monitoring instructions

Fairness

• Comparability statement — what makes scores comparable across candidates
• Structural requirements — opening → competency blocks → closing
• Difficulty calibration — target level-3 baseline, probe for 4-5

Your Key Deliverable

The complete adaptive rubric JSON — see Phase 9 Questions. This is the most critical science deliverable for unlocking adaptive mode.