Service Level Agreement

Short version. CourtFlow targets 99.5% monthly uptime for the signed-in dashboard and API on every active paid subscription, on a best-effort basis. We measure ourselves against the live status page at courtflow.ai/status. Court email ingestion, AI generation, and other features that depend on third-party providers are best-efforts and are not covered by an uptime number — the third-party dependency is too large for us to commit to a percentage we don't control. This is a best-effort target and is not backed by service credits.

1. Scope

This Service Level Agreement ("SLA") applies to the production CourtFlow AI service operated at courtflow.ai for an active, paid subscription. It does not apply to:

• The 14-day free trial;
• Public marketing pages (homepage, pricing, blog, compare, practice-area pages, legal pages);
• Public tools (free filing analyzer at /try, court rules reference at /rules, public probate intake portal, sales chat widget);
• Beta or preview features explicitly labelled as such in the application;
• Customer-controlled storage (Google Drive, OneDrive), customer-controlled email providers (Gmail, Outlook), and other third-party services CourtFlow integrates with. Those services have their own SLAs operated by their providers.

2. Uptime Target

CourtFlow targets 99.5% monthly uptime on a best-effort basis, measured as set out in §3, for the signed-in product surfaces at /dashboard and its sub-routes and the customer-facing API, on every active paid subscription. The target is best-effort and is not backed by a service-credit remedy. Other Service components — court-email ingestion cron jobs, AI document analysis, draft generation, and any other feature whose end-to-end completion depends on a third-party provider — are covered only by best-efforts and are not included in the uptime calculation.

3. How We Measure Uptime

Uptime in a calendar month is calculated as:

Downtime means a continuous period of at least five (5) minutes during which the dashboard or API returns 5xx errors for the majority of requests, or the dashboard fails to load for the majority of authenticated users, as confirmed by external synthetic monitors that probe the service from at least two geographic regions.

The public status page at courtflow.ai/status publishes the running uptime tally and the incident history that feeds the calculation. The status page is authoritative for SLA purposes.

4. Exclusions from Downtime

The following are not counted as downtime for SLA purposes:

• Scheduled maintenance. Planned maintenance windows announced at least 48 hours in advance via the status page and email to administrator addresses on file, performed outside U.S. business hours (Mon–Fri 9 AM – 6 PM ET) where reasonably possible. Aggregate scheduled maintenance will not exceed four (4) hours in any calendar month under normal circumstances.
• Emergency maintenance. Unplanned interventions necessary to address a security vulnerability, prevent imminent data loss, or comply with a legal requirement. Notice is provided as soon as practicable.
• Third-party-provider outages. Downtime caused by an outage at a Subprocessor (e.g., Vercel, Supabase, Google, Microsoft, Anthropic, Stripe, Resend, Upstash) where CourtFlow cannot route around the failure. We use multi-provider failover (Gemini model chain, multiple email providers) where architecturally feasible; where it is not, the underlying provider's outage flows through.
• Internet and last-mile. Downtime caused by general internet routing failures, the customer's ISP, the customer's VPN or proxy, or DNS resolver issues outside CourtFlow's control.
• Customer-side issues. Downtime caused by misconfiguration of the customer's account (e.g., revoked OAuth grants, expired refresh tokens), browser incompatibility, or use of features in a manner not supported by the documentation.
• Suspension for cause. Periods when access is suspended under the Acceptable Use Policy, the Refund & Cancellation Policy (failed-payment dunning), or for cause under the Terms of Service.
• Force majeure. Events beyond CourtFlow's reasonable control: natural disasters, war, civil unrest, government action, network or power outages at scale, and similar events.

5. Support Response Times

Support requests are triaged by severity as defined below. Response time is measured from receipt of a support email at support@courtflow.ai to the first substantive human reply, during U.S. business hours (Mon–Fri 9 AM – 6 PM ET).

Sev-1 and Sev-2 issues outside business hours are handled on a best-efforts basis. Security incidents reported under our Responsible Disclosure channel are triaged on a separate, 24/7 rotation regardless of subscription.

6. Changes to this SLA

CourtFlow may update this SLA from time to time. We will post the updated SLA at this URL, bump the "Last updated" date, and post a changelog entry. If a change materially reduces the commitments in §§2 or 5, we will give customers at least thirty (30) days' notice via the administrator email addresses on file before the change takes effect; the customer's sole remedy if it does not accept the change is to terminate the affected Service for cause within that notice period and receive a pro-rata refund of any prepaid fees covering the period after termination.

Report an outage

Current outage or degraded performance: check courtflow.ai/status, then email support@courtflow.ai.