Voice AI campaigns: configure a bounded run
Connect an eligible contact set, reviewed assistant, message template and execution limits before starting a campaign.
Table of Contents▼
A campaign coordinates multiple attempts; an assistant handles the conversation within an attempt. Keep these responsibilities separate when configuring outbound voice. A good prompt cannot repair an ineligible contact list, a missing route or an excessive retry policy.
Prepare the run
Record the purpose, eligible contact set, selected assistant version and intended outcome. Decide which actions are live and which test fixtures are used. Review contact permissions and suppression before importing real recipients; see contact controls.
Burki retains optional campaign services with configuration for templates and execution settings. Their presence is not a promise that every campaign/carrier/runtime combination has passed telephone acceptance.
Inspect templates and contacts
Preview the rendered greeting and agenda using representative records, including missing optional values and special characters. Do not place sensitive details in an opening greeting before the appropriate recipient is confirmed. A preview shows rendered text; it does not guarantee the full spoken conversation will follow it verbatim.
Use the contact preparation guide and template examples before execution.
Set explicit limits
Define total attempts, maximum concurrency, retry delay, permitted schedule and maximum financial exposure. Keep carrier and provider limits authoritative. There is no universal “safe” starting concurrency or retry count suitable for every account and use case.
A retry is a new potential contact and cost. It should not bypass newly applied suppression, a changed schedule or an unresolved previous action.
Verify a controlled attempt first
Use a number your team controls. Check two-way conversation, the actual action result, hangup, resource cleanup and settled usage. Then verify campaign-specific stop, pause and retry behavior within the authorized budget.
Review the run by outcome
Distinguish scheduled, attempted, answered, completed, failed and action-confirmed states. A completed call is not automatically a qualified lead or booking. Keep failed and pending operations visible for investigation.
Expand only when contact controls, conversation behavior and campaign execution have each been accepted. A high dial rate is not evidence of usefulness or compliance.
Ready to try Burki?
Create an assistant and check your available browser practice allowance.
Create your assistantTrial eligibility and available practice are shown in your workspace.