Operational Guarantees
Service Level Agreement
Our commitment to uptime, response velocity, and seamless timezone integration for enterprise partners.
1. Incident Response Times (TLA)
We categorize support requests and engineering blockers based on severity. For active production deployments under maintenance contracts, the following response times are strictly enforced:
| Severity Level | Definition | Target Response | Target Resolution |
|---|---|---|---|
| P1 (Critical) | Complete production outage. AI models failing, database down, or critical security vulnerability identified. | < 1 Hour | < 4 Hours |
| P2 (High) | Core feature degradation. App is online, but primary workflows (e.g. RAG generation, payment processing) are failing. | < 4 Hours | < 24 Hours |
| P3 (Medium) | Minor bug or UI defect that does not block core business functionality. | < 12 Hours | Next Sprint |
| P4 (Low) | General inquiries, feature requests, or non-critical infrastructure tweaks. | < 24 Hours | Scheduled |
2. Timezone Overlap & Communications
To eliminate the "offshore delay", Acadify Solution enforces a mandatory daily overlap window for all dedicated engineering teams working with US clients.
- 4-Hour PST/EST Overlap: Our tech leads and senior engineers are online during the first or last 4 hours of your working day for live pair programming, sprint planning, and PR reviews.
- Daily Loom Updates: Developers post 2-minute screen recordings at the end of their shift, visually demonstrating the code committed and features built.
- Slack Connect: We integrate directly into your company's Slack or Microsoft Teams workspace for real-time collaboration.
3. Infrastructure Uptime Guarantee
For environments hosted and managed by Acadify Solution (Cloud DevOps packages):
- We guarantee a 99.99% uptime SLA for production web servers and databases.
- Calculated monthly. If uptime falls below this threshold due to our managed infrastructure, service credits are automatically applied to the next billing cycle.
- Note: Excludes downtime caused by third-party API outages (e.g., OpenAI API drops) or scheduled maintenance windows (which are communicated 7 days in advance).