Service Level Agreement (SLA) - Zirvo

1. Service Commitment

Zirvo commits to employing best-in-class systems engineering reliability practices and maintaining its distributed, redundant, active-active multi-region architecture operating at a consistently high standard.

During the term of your formal license, Zirvo ensures that the In-Scope Service Components will be operationally online and available for consumption, respecting the Monthly Availability Index according to the plan contracted by the respective Tenant:

2. Availability Calculation

The monthly availability of the Zirvo Platform is calculated based on the following mathematical formula at the end of each calendar month in reference:

3. In-Scope Service Components

The guarantees described in this agreement apply strictly to the following Zirvo subsystems and infrastructure:

• Web Dashboard Platform: Accessibility to the web-hosted management panel by the customer.
• Public Integration APIs: Operational capacity to dispatch structured requests and API keys for request purposes.
• Verification Engine and Probes: The distributed core of workers that perform uptime, SSL, DNS, and latency readings per minute.
• Alert Processing and Ingestion: Internal dispatch and routing service for notifications from the Zirvo outbox to email and webhooks.

4. Exclusions

The following are explicitly excluded from any availability, indemnification, or monitoring commitments covered by this SLA:

• Monitored External Systems: All URLs, servers, and endpoints belonging to the Customer or third parties that Zirvo is monitoring.
• Customer Internet Provider Latency: Access impossibility generated by degraded connections at the customer's local operators.
• External DNS Resolution Errors: Failures in global DNS servers that pointually interfere with route resolution outside Zirvo's control.
• Third-Party Alert Destination Integrations: Notification delivery failures generated by systemic unavailabilities at Slack, Discord, Telegram, or PagerDuty endpoints.
• Improper User Configurations: Incorrect parameterization of HTTP headers, unexpected quotas, or timeouts more aggressive than the physical host can stably return.

5. Incident Classification

When an operational error is detected in our internal infrastructure, technical incidents are divided and handled under strict severity levels:

6. Incident Response Process

6.1. Incident Lifecycle

Zirvo's non-conformity handling is supported by constant automatic detection on our bus, followed by a triage flow, dedicated investigation by Site Reliability Engineering (SRE) teams, immediate logical isolation of shards or routes to contain damage, mitigation through operational canaries and surgical database resolution, followed by internal reports and full post-mortems.

6.2. Dynamic Transparency

In events classified under P1 or P2 criticality, the Zirvo team will maintain periodic active updates on the operational status panel and notify the linked administrative contact channels of tenants clearly, ensuring full transparency.

7. Maintenance Windows

7.1. Scheduled Maintenance

Zirvo periodically runs structured maintenance routines to apply security patches to PostgreSQL, optimize indexes in metrics databases (Timescale), and improve network components. These maintenance windows are preferentially scheduled during low-traffic hours (between 00:00 and 05:00 Brasilia time).

7.2. Communication and Downtime Exemption

When communicated to Customers via administrative email or internal platform bulletins with a minimum of 48 business hours advance notice, the hours elapsed during Scheduled Maintenance Windows will not be considered, for any legal or technical purpose, as systemic Downtime, and will not factor into credit or compensation equations.

8. Service Credits

8.1. Compensation Model

If Zirvo fails to meet the guaranteed Monthly Availability Index for a given corporate billing cycle, the Customer will be entitled to Service Credits applicable as a direct deduction on the subsequent monthly invoice:

8.2. Credit Request Process

Compensation is not generated automatically. To claim the benefit, the qualified billing representative of your company must submit a substantiated request to legal@zirvo.com.br within 30 business days, attaching panel logs, the incident date, and operational evidence. Zirvo will technically analyze the official internal metric and return a conclusive assessment within 10 consecutive business days.

9. Monitoring and Measurement

9.1. Official Provider Metrics

To settle technical disputes about instability assessments between networks, the measurements and audit logs natively and internally provided by Zirvo's servers and buses prevail as the sovereign technical source of truth, provided the collection accuracy is demonstrated in internal reports and logs. Discrepancies based exclusively on monitoring tools installed on the customer's own hosts or other channels without technical isolation are not accepted as official evidence.

10. Customer Responsibilities

To ensure monitoring performance is maintained intact and without bottlenecks, customer organizations fully agree to:

• Ensure correct management of team credentials (seats) and authentication (API Keys);
• Avoid improper conduct based on overloading, unauthorized stress testing, and computational concurrency abuses prohibited by our AUP;
• Configure intervals and alerts with technical parameters appropriate to the actual physical delivery capabilities of their respective hosting servers.

11. Monitoring Disclaimer

The activity performed by Zirvo is limited to the scientific provision of telemetry data, network port observability, and dispatch of reports and notifications in "As-Is" format.

12. Force Majeure

Zirvo will not be considered in default or liable for technical failures and availability deviations generated by unforeseeable events that exceed the reasonable capabilities of its direct governance. Included in this clause are physical breaks in global underwater network cables, natural disasters from extreme weather, declared civil wars, blackouts and massive structural failures in public utility power grids, virtual terrorist acts, sovereign government legal interventions, and global electronic catastrophes in the main infrastructure of public cloud providers (AWS, GCP, or Azure) of multi-region scope.

13. SLA Changes

We reserve the legitimate right to make minor changes or revise the technical-legal wording expressed in this SLA in order to accommodate computational advances and regulatory requirements. Should extensive structural modifications occur that could reduce the current guarantees of existing business plans, affected tenants will be notified with a minimum of 30 days advance notice via official administrative email, with the update implemented only at the turn of the contracted regular billing cycle.

14. Enterprise Support

For customers on the High-Scale Enterprise Plan, support is operated through direct and escalated service channels for rapid resolution of operational bottlenecks: