RedHub AI
How the gate works
How the Voice-Agent Conversation QA & Escalation gate works
This tool grades one voice-agent call across nine craft and trust signals - the shipped six-call sample. Each signal is scored 0–2; the verdict is the weakest one — then a gate that can only make it worse.
The 9 signals
The gate compares two readings and takes the worse: the craft-derived band (the weakest assessed mark across task resolution, the four applicable policy and disclosure controls, conversation handling and tone) and the graduated Trust gate on factual grounding and escalation correctness. A trust mark of 1 caps the final result at SOFT-HANDLED; a trust mark of 0 forces MISHANDLED. The gate can worsen—but never improve—the craft-derived band. Separately, a policy and disclosure control marked Not assessed prevents RESOLVED: a call that would otherwise read RESOLVED reads SOFT-HANDLED.
Why: Task resolution, the four applicable policy and disclosure controls, conversation handling, and tone are all marked 2 - the craft-derived band is RESOLVED AS DESCRIBED. But factual grounding is 0: the bot stated an unverified account balance. A trust mark of 0 triggers the graduated Trust gate and forces the final result to MISHANDLED. The policy-and-disclosure mark reflects the reviewer’s assessment of the transcript and supplied context. It is not a determination that the call complied with the TCPA, recording-consent laws, telemarketing rules or other applicable requirements.
A clean sheet didn’t save it. The verdict is the weakest signal — and a top trigger is dispositive on its own — so where a signal is short, the fix first row names it. Grades calls that already happened, from your own marks. Connects to no live agent, and it never scores people.