Service Level Agreement
Effective date: 21 February 2026 · Applies to all BatchCortex subscription tiers
1. Uptime Commitment
| Tier | Monthly Uptime Target | Measurement |
|---|---|---|
| Monitor | 99.5% | Dashboard + API availability |
| Analyze | 99.5% | Dashboard + API + report generation |
| Guard | 99.9% | All services + escalation system |
Uptime is measured monthly excluding scheduled maintenance windows.
2. What Counts as Downtime
- Dashboard inaccessible for >5 consecutive minutes
- API returning errors on >50% of requests for >5 minutes
- Escalation system failing to send alerts for confirmed anomalies
- Report generation unavailable for >30 minutes
3. What Does Not Count as Downtime
- Scheduled maintenance (minimum 48 hours notice via email)
- Force majeure events
- Issues caused by customer's on-site network or OPC-UA infrastructure
- Edge agent connectivity issues on customer hardware
- Supabase, Vercel, or Mistral AI platform-wide incidents
4. Edge Agent Availability
The edge agent runs on customer hardware inside the customer network. BatchCortex is not responsible for edge agent uptime — this is the customer's infrastructure. The store-and-forward buffer ensures no data is lost during cloud connectivity interruptions of up to 72 hours.
5. Service Credits
If monthly uptime falls below the committed target:
| Actual Uptime | Credit |
|---|---|
| 99.0% – 99.5% | 10% of monthly batch fees |
| 98.0% – 99.0% | 25% of monthly batch fees |
| Below 98.0% | 50% of monthly batch fees |
Credits are applied to the following month's invoice. Credits are the sole remedy for downtime — BatchCortex liability is limited to fees paid in the preceding 12 months.
6. Scheduled Maintenance
- Standard window: Sundays 02:00–04:00 CET
- Minimum notice: 48 hours via email
- Emergency maintenance: As required with maximum reasonable notice
7. Support Response Times
| Tier | Critical Issue | Standard Issue |
|---|---|---|
| Monitor | 8 business hours | 48 hours |
| Analyze | 4 business hours | 24 hours |
| Guard | 4 hours (including weekends) | 12 hours |
Critical = escalation system down, dashboard inaccessible, data integrity concern.
8. Contact
For SLA-related inquiries, contact: vilmer@batchcortex.com