Starter ($49/mo)
99.5% monthly uptime target
Credit policy: Service credits not bundled. Best-effort SLA with email-channel support.
Monitoring: 5-minute synthetic monitoring; status page real-time.
Uptime SLA
Uptime targets, history and incident response
Published SLA targets per tier, 6-month historical uptime metrics, service-credit policy and a 6-step incident-response procedure. Real-time status: status.slatech.ai. Pairs with security controls, architecture overview and compliance glossary.
SLA targets per tier
Pro ($249/mo)
99.9% monthly uptime target
Credit policy: 10% monthly fee credit if uptime drops below 99.5%. 25% if below 99%.
Monitoring: 5-minute synthetic monitoring + per-tenant query latency tracking; status page real-time.
Enterprise ($499/mo)
99.9% monthly uptime target
Credit policy: 10% monthly fee credit if uptime drops below 99.5%. 25% if below 99%. 50% if below 95%.
Monitoring: 1-minute synthetic monitoring + per-tenant query latency tracking; priority status notification.
Custom Solutions (custom)
99.95% monthly uptime target (single-tenant deployment)
Credit policy: Negotiated in the MSA. Typical: 15% monthly credit per 0.1% under target. Direct CSM escalation channel.
Monitoring: 15-second synthetic monitoring; dedicated Pingdom dashboard; on-call SRE escalation procedure.
Historical uptime (last 6 months)
| Month | Uptime | Incidents | Notes |
|---|---|---|---|
| June 2026 (m-t-d) | 99.97% | 0 | No production incidents. WAF blocked 2.1M malicious requests. |
| May 2026 | 99.94% | 1 | 12-minute degradation 2026-05-17 due to a vector-store index-rebuild - resolved automatically. |
| April 2026 | 99.99% | 0 | No production incidents. Strong month. |
| March 2026 | 99.92% | 1 | 8-minute incident at our CDN / edge-security provider affecting global edge routing - out of SLAtech direct control but still counted. |
| February 2026 | 99.96% | 0 | Brief cloud-region partition tested failover - no customer impact. |
| January 2026 | 99.98% | 0 | Stable month. Wedge upgrade to .NET 10 was zero-downtime. |
Rolling 6-month average: 99.96%. Real-time live data: status.slatech.ai.
6-step incident-response procedure
-
1. Detection
Error-monitoring alert or synthetic-monitoring failure triggers PagerDuty-equivalent rotation. Acknowledged within 5 minutes for P1, 15 minutes for P2.
-
2. Triage
On-call SRE assesses impact scope (which tenants, which features). Status page updated within 10 minutes of acknowledgment with initial impact assessment.
-
3. Mitigation
Standard runbook applied (failover, rollback, hotfix). Median mitigation time for known patterns: 15 minutes. Novel incidents: 30-60 minutes.
-
4. Resolution
Service confirmed restored across synthetic monitors and a sample of tenant transactions. Status page updated with resolution timestamp.
-
5. Post-mortem
Public post-mortem published within 5 business days. Includes timeline, root cause, customer impact (number of tenants affected, duration), corrective actions (technical + process).
-
6. Customer follow-up
Affected Pro/Enterprise/Custom Solutions tenants receive direct email with incident details + applicable service credit. Custom Solutions tenants get a dedicated post-mortem call.
Reporting a security or uptime incident?
SECURITY incidents: security@slatech.ai. UPTIME / customer impact: status.slatech.ai live feed.
Buyer evaluation tools
Four self-serve tools for evaluating SLAtech (or any AI chatbot vendor) without a sales call:
Frequently asked questions
What's SLAtech's uptime SLA?
Starter 99.5% monthly. Pro and Enterprise 99.9% monthly. Custom Solutions 99.95% monthly (single-tenant). Service credits apply per the tier-specific credit policy. Status real-time at status.slatech.ai.
What's the historical uptime over the last 6 months?
January 99.98%, February 99.96%, March 99.92%, April 99.99%, May 99.94%, June (month-to-date) 99.97%. Rolling 6-month average: 99.96%. Two minor incidents (1 vector-store index-rebuild, 1 CDN-edge partition at our edge provider) within that period.
How is uptime measured?
Synthetic-monitoring transactions from multiple cloud regions hitting /v1/query/ask with a known query, every 5 minutes (Starter/Pro), every minute (Enterprise), every 15 seconds (Custom Solutions). Includes API latency budget (response < 5 seconds = healthy). Independent verification available via a status-page widget.
How are incidents reported and resolved?
6-step procedure: detection (error monitoring / synthetic) → triage (status page within 10 min) → mitigation (runbook applied) → resolution → public post-mortem within 5 business days → customer follow-up with service credit. Enterprise tenants get a dedicated post-mortem call.
Is there real-time uptime dashboard?
It shows whether each of twelve production components is answering right now - the API, the admin platform, the main site and the nine vertical hubs. A Worker job probes them and the page caches the result for a minute. It is an availability board, not a metrics dashboard: latency and delivery rates live in each tenant's own admin analytics.
What happens during a security incident?
Same procedure plus immediate notification on our security-incident channel, optional engagement of legal counsel (privacy lawyers), and 24-hour customer notification SLA for significant incidents (per NIS2 alignment). Affected data, scope and mitigation enumerated in the post-mortem.