Service Level Agreement (SLA)

GliminTor OS Service Level Agreement (SLA) Uptime commitments, incident response, and service remedies Product GliminTor OS Operator XBLANC SPÓŁKA Z OGRANICZONĄ ODPOWIEDZIALNOŚCIĄ Support contact [email protected] Effective date April 2026 *v3.0 — April 2026*

This SLA applies to the GliminTor OS web application (app.glimintor.com) and API (api.glimintor.com). It does not cover third-party connected platforms, AI providers (OpenRouter, fal.ai), Zoom infrastructure, or planned maintenance windows.

Maker 99.0% monthly uptime Creator 99.0% monthly uptime Pro 99.5% monthly uptime Studio 99.5% monthly uptime Agency 99.9% monthly uptime Enterprise 99.9% monthly uptime (or custom per contract) "Monthly uptime" = (total minutes − downtime minutes) / total minutes × 100. "Downtime" = consecutive unavailability of 5+ minutes, excluding excluded periods.

• Scheduled maintenance (48+ hours notice via in-app banner or email). • Emergency maintenance to protect security or data integrity. • Outages caused by third-party providers (Railway, Cloudflare, OpenRouter, fal.ai, Stripe, Clerk, Zoom). • Force majeure events. • Customer-caused outages (misconfiguration, exceeding API limits).

P1 — Critical (Service down) Response: 30 min. Resolution target: 4 hours. P2 — Major (Core feature unavailable) Response: 2 hours. Resolution target: 8 hours. P3 — Minor (Degraded performance) Response: 8 hours. Resolution target: 48 hours. P4 — Low (Cosmetic / non-critical) Response: 24 hours. Resolution target: 14 days. Priority Support subscribers (€99/mo) receive a dedicated support channel with guaranteed 4-hour first-response for P1–P2 incidents.

5.1 Video Generation Queue: • Video generation jobs are lowest priority in the processing queue behind LLM text and image tasks. • Generation failures caused by fal.ai outages are retried automatically up to 3 times. Credits are refunded if all retries fail. • Video queue delays during high platform load do not constitute a breach of this SLA. 5.2 Broadcast: • Broadcast message delivery depends on third-party platform APIs (WhatsApp, Telegram, Meta, Messenger). Xblanc does not guarantee delivery timelines for messages pending platform-side delivery. • Broadcast queue processing is best-effort and does not carry a delivery time SLA. • Broadcast access may be suspended without SLA credit implications if spam or policy violations are detected. 5.3 Auto-Engage: • Auto-Engage response generation is subject to the same LLM queue as other AI text operations. • Response dispatch depends on third-party platform API availability and rate limits. 5.4 Live Studio: • Live Studio streaming depends on destination-platform RTMP endpoint availability (Facebook Live, YouTube Live, Twitch, etc.). Destination-platform outages are excluded from uptime calculations.

99.0%–99.4% uptime (vs target) 5% of monthly subscription as credit 98.0%–98.9% uptime 10% of monthly subscription as credit 95.0%–97.9% uptime 20% of monthly subscription as credit Below 95.0% uptime 30% of monthly subscription as credit SLA credits applied to next invoice. Not redeemable for cash. Claims must be submitted within 30 days to [email protected]. Maximum credit per calendar month: 30% of monthly subscription fee.

• PostgreSQL: automated daily backups, 7-day retention (Railway). • Object storage: replicated across Cloudflare R2 infrastructure. • Recovery Point Objective (RPO): 24 hours. • Recovery Time Objective (RTO): 4 hours for P1 incidents.

Scheduled maintenance: typically 02:00–05:00 CET/CEST on Sundays. Notification at least 48 hours in advance via in-app banner and email to workspace administrator.

SLA credits are the sole remedy for uptime failures. This SLA does not apply to free trials or sandbox environments.

Incidents and support: [email protected] SLA credit claims: [email protected] Postal: Xblanc Sp. z o.o., ul. Henryka Sienkiewicza 22, 60-818 Poznań, Poland