Calling a list is easy; operating a dialer is not
A naive outbound system can iterate through phone numbers and place calls. A production system has to control how quickly those calls start, what happens when they fail, how retries are scheduled, how many sessions can run at once, and when the campaign must stop.
Pacing is a capacity decision
Cally’s background campaign worker can enforce dialed calls per minute and dispatch leases. This keeps concurrency within defined limits and prevents multiple workers from unintentionally processing the same job. The pacing rate should also reflect the capacity of connected telephony, AI inference, and human transfer queues.
Retries need policy
Busy, unanswered, and failed calls should not immediately return to the front of the queue. Cally supports retry delays and daily-attempt limits. A thoughtful campaign distinguishes retryable outcomes from terminal outcomes and respects a contact’s local calling window before another attempt is made.
Budgets are operational guardrails
Automated calling consumes carrier minutes and AI usage. Cally can enforce campaign budget cutoffs and also prevent outbound runs when the account balance falls below defined thresholds. This makes spend a runtime control rather than something discovered after the campaign ends.
Reliability continues after disconnect
The feature sheet describes a durable result outbox that writes call results locally before delivering them to the cloud. That pattern protects summaries and recordings from being lost if connectivity drops at the wrong moment. A campaign is not complete when the phone hangs up; it is complete when the outcome is safely recorded.
Model your campaign as a queueing system: arrival rate, active capacity, retry policy, human overflow, and budget. The conversation is only one stage in that system.