Resources · Articles
Articles

Notes from the floor on governed voice AI.

Short, practical pieces for contact center leaders who have to make AI work on regulated calls. No hype, no borrowed statistics, just how we think about the problems that decide whether a program reaches production.

Latest

Three articles to start with.

September 22, 2026

From sampling to every call: keeping humans in calibration

Scoring every conversation changes what QA can see. It only works if the scorecard stays yours and your QA leads keep the machine honest.

Read →
September 15, 2026

What a good AI to human handover looks like

The moment a call moves from an AI agent to a person is where trust is won or lost. Context is what makes the difference.

Read →
September 8, 2026

Why consequential actions need a write gate on AI calls

Let the AI talk at machine speed. Make it stop before it changes money, an account or a promise. Here is why, and how it works.

Read →
What we write about

The parts of voice AI that vendors skip in the demo.

Approvals, handovers, scoring, audit and the rules that apply on every call. These are the things your compliance and operations teams will ask about, so we write about them first.

  • GovernanceWrite gates, audit chains and who signs off what.
  • OperationsHandovers, coaching and the floor.
  • QualityScorecards, calibration and disputes.
spine·voice // auto qa · scorecard v7call #40316_
Auto QA · Member servicesample
88score
0violations
5:03handle
+0.2sentiment
CriterionResultWeight
Identity verified before disclosure✓ pass
Summarized next stepspartial
1 item sent to QA leadDispute
Auto QA scorecard, sample data · rendered by spine·voice
Book a demo

See the ideas running on a live call.

Pick any article and we will show you the part of the platform it describes, working end to end.