// loading
The OpsTicket Service Level Agreement applies to all paid subscriptions (Recruit, Enterprise Suite, Enterprise+). Free candidate accounts are governed by the Terms of Service.
"Uptime" means the percentage of total minutes in a calendar month during which the OpsTicket production application (app.opsticket.com, candidate assessments, recruiter pipeline, API) is reachable and returns a 2xx or 3xx status to a synthetic monitor we run.
Uptime tier varies by plan (see the per-tier matrix below). Downtime is the sum of consecutive failed checks across our monitoring locations.
Excluded: scheduled maintenance announced 7 days in advance, downtime caused by customer misconfiguration, force majeure (see operational fine print), and third-party identity providers (SSO outages on Okta, Azure AD, etc. are tracked separately).
| monthly uptime | service credit | maximum credit | how to claim |
|---|---|---|---|
| greater than or equal to target | · none | · | · |
| < target (above 99.50%) | 10% of monthly fee | 10% | auto-applied on next invoice |
| < 99.50% | 25% of monthly fee | 25% | auto-applied · we email you |
| < 99.00% | 50% of monthly fee | 50% | auto-applied · we email you |
A P0 incident is a total outage of the production application affecting all paying customers, or a confirmed data exposure incident. We acknowledge P0 reports within 1 hour during business hours (09:00 to 18:00 ET, Mon to Fri) and begin remediation within 4 hours.
Status page: status.tryopsticket.com publishes an incident timeline within 30 minutes of acknowledgment. Post-mortem published within 5 business days for any P0.
Outside business hours, the on-call engineer responds within 4 hours for Enterprise tiers and within 1 hour for Enterprise+.
For 30 days after you cancel or we terminate your subscription, you retain read-only access to export: candidate sessions, replay artifacts (keystroke logs, transcripts, scoring rubrics), recruiter pipeline state, and posting history. Enterprise tiers receive 60 days; Enterprise+ receives 90 days.
Export formats: CSV (sessions, candidates, postings), JSON (full event stream including replay events), MP4 (per-session replay rendering on request).
After the export window, data is purged from active systems within 7 days and from cold backups within 90 days. Purge is verifiable on request.
For Enterprise and Enterprise+ tenants, the worst-case data loss if we lose the primary database is the last 5 minutes of writes. Postgres (Supabase managed) runs point-in-time recovery; logical backups are continuous.
Recovery Time Objective (RTO) is 60 minutes for full production restore on Enterprise and 30 minutes on Enterprise+. We rehearse failover on a quarterly cadence.
If a candidate has an in-progress assessment and we publish a rubric update to that scenario, the in-progress session continues to be scored against the rubric in effect when they started. Their session is never re-scored against a newer rubric without their explicit replay submission.
Rubric versions are pinned at session start, recorded in the session metadata, and visible to recruiters in the replay header. This is the same trust contract we publish in the candidate Terms.
If we miss the number, the credit appears on next month's invoice automatically · we don't make you fight for it, and we don't make you ask.
All paying customers affected, or candidate data confirmed reachable to an unauthorized party. Status page incident opened. Senior engineer plus founder paged.
Cannot create or grade assessments, recruiter compare view down, SSO login broken. Workaround may exist but is unacceptable for production hiring.
Non-critical feature broken, slow page loads, intermittent UI bug. Workaround exists. Production hiring continues.
Visual bug, typo, low-traffic page error, documentation gap.
Routine maintenance is scheduled within this window when it must cause user-visible downtime. We announce the specific Saturday 7 days in advance on status.tryopsticket.com and via email to listed admins.
Emergency maintenance (security patches with active CVE exploits) may occur without notice. We post within 15 minutes and write up within 24 hours.
Replay artifacts and session transcripts are stored in managed object storage with provider-side replication. Postgres (Supabase managed) runs point-in-time recovery; daily snapshots are retained for 35 days.
Restore testing runs quarterly.
SLA does not apply to: upstream cloud-provider regional outages affecting multiple providers, BGP route incidents, undersea cable cuts, government-imposed shutdowns, or natural disasters preventing access to data centers.
We do still owe you a post-mortem and a credible mitigation plan.
SLA claims, contract redlines, procurement questionnaires, security review requests: a real human replies within one business day. Counter-signed master SLA available for Enterprise tiers on request.
Procurement compliance note: OpsTicket is operated by IT Custom Solution LLC (d/b/a OpsTicket), 420 Lexington Avenue, Suite 1402, POB 1005, New York, NY 10170. We are not currently audited under any AICPA SSAE-18 framework and do not represent that an audit is in progress. We hold no current GSA Schedule (pursuing GSA MAS). 8(a) Application Submitted 2026. For full security posture see /security.