AI at the core
AI that investigates — not just summarises.
Summarising is easy and mostly useless — it tells you what you already have. These four engines do the work you would otherwise assign to an analyst, and every answer stays attached to the evidence that produced it.
Engine 01
Cxorbi Analyst
Ask the question you would have asked an analyst.
Runs its analysis tools across your behavioural, replay and feedback data, then returns a decision brief with every figure linked to the evidence behind it.
What this name refers to: Agentic tool-calling investigator over the product's read-only analysis tools.
Engine 02
FrictionSense
Struggle does not announce itself.
Detects rage clicks, dead clicks, looping and thrash automatically, then ranks sessions Smooth through High so the worst surface first.
What this name refers to: Friction signal detection and session scoring, web and mobile.
Engine 03
VoiceLens
Nobody reads four thousand reviews.
Turns open text into sentiment, themes, intent and urgency — across survey responses and public reviews alike, in the languages your customers actually write in.
What this name refers to: Text analytics and sentiment provider services behind Ask and Listen.
Engine 04
Evidence Link
A number you cannot check is an opinion.
Keeps every figure traceable to the sessions and responses that produced it, so an answer can always be opened and verified rather than trusted.
What this name refers to: Session-id joins across replay, funnels, occurrences and survey responses.