OPERATIONAL GUARANTEES

Service Level Agreement

Our commitment to uptime parameters, response thresholds, and structured timezone integration for enterprise development partners.

1. Incident Response & Resolution

Support tickets and technical incidents are categorized strictly by severity levels. For deployments under active enterprise support agreements, the following response boundaries are guaranteed:

Severity Level Operational Definition Target Response Target Resolution
P1 (Critical) Complete system outage. Production endpoints failing, core databases unresponsive, or active security exploit identified. < 1 Hour < 4 Hours
P2 (High) Significant feature degradation. App is online, but primary APIs (e.g., RAG vector generation, payment gateways) are failing. < 4 Hours < 24 Hours
P3 (Medium) Minor functional defect or non-blocking UI anomaly that does not impede primary business workflows. < 12 Hours Next Release Cycle
P4 (Low) General technical queries, minor feature inquiries, or non-critical cloud configuration optimizations. < 24 Hours As Scheduled

2. Timezone Overlap & Engineering Sync

To optimize collaboration and eliminate latency in communication, Acadify enforces structured daily working overlaps for all dedicated systems teams.

4-Hour Active Window

Our tech leads and systems engineers maintain a minimum of 4 hours daily timezone overlap matching your core working hours (PST, EST, or GMT) to conduct live integrations, planning sessions, and system reviews.

Asynchronous Briefs

All developers submit video recordings at the end of each shift, documenting committed codebase blocks, test validations, and current sprint impediments to ensure visibility.

3. Uptime Guarantees & Service Credits

For production environments hosted and managed under Acadify's Cloud DevOps agreements, we guarantee a 99.99% Monthly Uptime Percentage.

Monthly Uptime Level Service Credit applied to invoice
< 99.99% but ≥ 99.0% 10% Credit
< 99.0% but ≥ 95.0% 25% Credit
< 95.0% 100% Credit
Exclusions & Scope Limits

Service guarantees do not cover downtime resulting from third-party foundational models (e.g., Anthropic Claude or OpenAI API networks), client-controlled infrastructure, or scheduled maintenance windows communicated at least 7 days in advance.