> ## Documentation Index
> Fetch the complete documentation index at: https://developers.callaro.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Service Levels and Rate Limits

> Document availability expectations, maintenance posture, and how to build around published limits.

# Callaro aims for predictable availability and controlled maintenance windows.

Customers should align on service expectations, incident communication, and integration-side resilience before production launch.

## Availability and communication model

* Status communication endpoint: `https://status.callaro.ai`
* Incident updates: published during active service impact windows
* Planned maintenance: announced in advance for customer-impacting changes

## Customer integration expectations

<Steps>
  <Step title="Implement retry and timeout policy">
    Configure exponential backoff for transient failures and keep strict timeout ceilings in all automation workers.
  </Step>

  <Step title="Design idempotent write paths">
    Billing, CRM updates, and webhook consumers should tolerate duplicate delivery and replay safely.
  </Step>

  <Step title="Monitor endpoint-family health">
    Separate dashboarding for auth failures, webhook failures, and call execution API failures.
  </Step>

  <Step title="Define escalation ownership">
    Keep a named primary owner and backup owner for incident response in your integration team.
  </Step>
</Steps>

## Support and escalation workflow

| Severity | Typical trigger                                    | Expected customer action                                                  |
| -------- | -------------------------------------------------- | ------------------------------------------------------------------------- |
| Critical | Major call execution or API unavailability         | Initiate incident bridge, pause non-essential writes, monitor status page |
| High     | Partial degradation (for example delayed webhooks) | Enable fallback/retry mode, notify operators                              |
| Medium   | Isolated endpoint issues                           | Queue retries and continue monitored execution                            |

<Note>
  SLA and support commitments may vary by contract tier. Use your enterprise agreement as the governing document for formal service guarantees.
</Note>