Houston AI

Model Help · Feb 28–Mar 30, 2026 · ET

How this works

A technical explainer of how Houston scores message interactions.

  • Each outbound creates a reply window.
  • Replies and outcomes are attributed within that window.
  • Metrics are computed per lead per day.
Limits
Calls are logged as outcomes; initiator may be unknown.
What to do next
Use this to audit scoring logic for a template.

A technical walkthrough of how the model computes exposure, replies, negative reactions, silence, and task completion for a selected template.

Select a template

Loads explainer below
Tip: start with a high-exposure template from the Action Queue for the clearest signal.

Template Explainer

Technical view

Select a template to load the explainer.

Representative snippet