Communications infrastructure

Twilio routing, queue reliability, and voice/SMS safety controls

Technical operations for outbound and inbound messaging, queue recovery, voice fallback, and provider-dependent incident response.

Outbound channels

SMS + Voice

All critical workflows can pivot between text and call paths when the primary channel degrades.

Queue retry budget

3 attempts

Retries are capped and spaced to preserve carrier reputation and avoid duplicate caregiver contact.

Voice fallback trigger

6 min

Calls are armed when the member or guardian path crosses the configured no-response threshold.

Provider observability

Full stack

Queue state, Twilio callbacks, status webhooks, and escalation side effects are tracked together.

Operational briefing

The context operators need before making manual decisions.

Queue management policy

Operators can run, pause, retry, and remove work items, but destructive actions remain senior-only because queue state affects both safety posture and audit integrity.

Voice fallback principles

Voice is used deliberately when timing, urgency, or repeated delivery failure suggests SMS is no longer sufficient, and every transition should preserve a clear audit trail.

Provider fault containment

If Twilio status callbacks drift, the platform should freeze risky automation, preserve visible exceptions, and give operators a deterministic recovery path instead of partial retries.

Operator checklist

Core reviews to complete during steady-state operation.

Monitor failed or paused queue items before they accumulate into stale outreach.

Validate webhook callback health for delivery, recording, status, and inbound response routes.

Check voice fallback campaigns for duplicate attempts and timing collisions.

Review carrier or number-level issues when failures cluster by destination prefix.

Document every manual queue removal that changes the expected outbound narrative.

Initiatives and ownership

Current operating priorities with named ownership and review rhythm.

Failure clustering

Group failed sends by carrier, geography, and workflow so the team can see whether an outage is local, systemic, or account-specific.

Owner: Platform reliability

Cadence: Continuous

Voice studio governance

Keep prompts, fallback scripts, and monitor transfer paths versioned so production call flows remain explainable after changes.

Owner: Communications engineering

Cadence: Release based

Manual override audit

Every run-now or retry action should be attributable to a person, a reason, and the queue context that justified intervention.

Owner: Operations security

Cadence: Weekly audit

Reference material

Documentation designed to answer follow-up questions without leaving this section.

Twilio callback dependency map

Lists which platform behaviors depend on delivery, status, recording, and inbound webhooks and what degrades when they stop arriving.

Queue recovery runbook

Step-by-step instructions for pausing dispatch, replaying safe work, and reconciling stale queue items after provider incidents.

Voice script catalog

Current prompt inventory for member calls, guardian escalation, confirmation flows, and monitored call transfers.