KKOS does not hand you a personality label.
The whole loop is designed to be inspected and corrected — not to hand down a verdict.
Pick camera, voice or text. The AI opens the conversation and follows your lead — no script, no form.
Behavioural signals accumulate across the 11 dimensions while the session runs, weighted by evidence strength.
The engine returns a structured model with evidence, confidence, uncertainty and alternative readings.
Disagree with a trait? Mark it. The model re-weights and regenerates — your view is part of the data.
Signal intensity across the 11 dimensions over conversation time. Brighter cells mean stronger evidence for a trait at that moment — the same heat surface the engine reads before it writes the model.
Communication style
How does this person build a sentence — do they lead with the point, or arrive at it?
Directness
Do they say the thing, or talk around the thing?
Openness
How much do they volunteer before anyone asks?
Engagement
Do they lean in, or keep a polite distance?
Decision style
Do they decide from evidence, from gut, or from what other people expect?
Uncertainty response
What happens when they do not know?
Reflection style
Do they turn a question back on themselves?
Conflict style
Do they push back, or route around friction?
Trust and disclosure
At what point do they say something real about themselves?
Emotional expression
How much of what they feel actually reaches the words?
AI interaction style
Do they treat the AI as a tool, a stranger, or a colleague?
These are the questions each dimension is designed to answer — not scores, and not measurements of you. Every one of them has to be supported by evidence quoted from the session before it appears in your model.
The visual modes below are animated illustrations.