Overview
This Service Level Agreement ("SLA") describes Tensor Machine's uptime commitments and remedies for service unavailability. This SLA applies to paid accounts. Free tier accounts are not covered by uptime commitments but receive commercially reasonable availability on a best-effort basis.
Uptime Commitment
Tensor Machine commits to the following monthly uptime targets for the API inference service:
Growth plan: 99.5% monthly uptime.
Pro plan: 99.9% monthly uptime.
Enterprise plan: 99.95% monthly uptime.
Monthly uptime is calculated as: ((total minutes in month − downtime minutes) / total minutes in month) × 100. Downtime means periods where the API inference endpoint returns 5xx errors for more than 50% of requests in a 5-minute rolling window, as measured by Tensor Machine's monitoring infrastructure.
Exclusions
The following are excluded from uptime calculations:
Scheduled maintenance windows, announced at least 48 hours in advance via the status page (status.tensormachine.ai) and email notification. Events caused by factors outside Tensor Machine's reasonable control (force majeure: natural disasters, power grid failures, acts of government). Outages caused by Customer's own infrastructure, code, or API usage patterns. Issues with third-party networks or services not directly controlled by Tensor Machine. Periods of suspension under the Terms of Service or AUP.
Service Credits
If monthly uptime falls below the committed level, eligible Customers may claim service credits:
99.0% – below commitment: 10% credit of monthly spend.
95.0% – 98.9%: 25% credit of monthly spend.
Below 95.0%: 50% credit of monthly spend.
Credits are applied to the next billing period. Credits are not redeemable for cash. Credits are the sole and exclusive remedy for SLA failures. To claim a credit, submit a request to sla@tensormachine.ai within 30 days of the incident, including the incident date, duration, and impact on your workload.
Support Response Times
Support is provided via email at support@tensormachine.ai and in-console ticketing. Response time commitments by plan:
Free tier: Community forum + documentation only. No guaranteed response time.
Growth plan: First response within 2 business days for standard issues. 1 business day for billing and account issues.
Pro plan: First response within 8 business hours for standard issues. 4 business hours for critical production issues (P1). Business hours: 09:00–18:00 IST, Monday–Friday.
Enterprise plan: First response within 2 hours for critical (P1) issues, 24×7. 4 business hours for high (P2) issues. Dedicated account manager and Slack channel available on request. Quarterly service reviews.
Priority classifications: P1 — complete service outage or severe degradation impacting production. P2 — significant degradation or partial outage. P3 — moderate impact with workaround available. P4 — general questions and feature requests.
Status and Incident Communication
Real-time service status is available at status.tensormachine.ai. Subscribe to status updates to receive email or webhook notifications for incidents and maintenance. Post-incident reports (PIRs) are published within 5 business days of resolution for P1 incidents affecting multiple customers.
Changes to This SLA
Tensor Machine may update this SLA from time to time. Upgrades to commitments take effect immediately. Reductions to commitments will be communicated 30 days in advance and will not apply to any period already covered by a pre-paid subscription.