24/7/365 Managed Service
We run hotel commerce. You focus on guests. White-glove operations, 99.95% uptime SLA, and live agents included.
Fully managed. Zero IT overhead.
Software vs. Service
Most hotel technology is software you install and manage. Hotel commerce is a service we deliver—operations, monitoring, and live agents included.
Traditional Hotel Software
CordiantOne (Managed Service)
Result: Hotel commerce without IT overhead. You define your brand voice. We handle everything else.
Live agent support included
98% of interactions: autonomous. 2% that need human judgment: our brand-trained agents step in seamlessly. Guest never knows. You never get the call.
Brand-Trained Agents, 24/7/365
When complexity requires a human, our live agents—trained on your brand voice, policies, and SOPs—take over with full conversation context.
Your staff: Not on call. Not interrupted at 2am.
Our agents: Handle it seamlessly. Included in managed service.
Availability commitments
SLA measurements calculated monthly. Scheduled maintenance announced in advance and excluded from calculations.
Uptime & Response SLAs
CordiantOne commits to the following availability and response time targets across all service components.
| Service Component | Monthly Commitment | Notes |
|---|---|---|
| Hotel commerce platform (Guest-facing & backend) |
99.95% | ~22 minutes max downtime/month. Excludes scheduled maintenance & third-party infrastructure outages. |
| AI conversation layer (Automatic multi-LLM failover) |
99.9% | Multi-provider redundancy with transparent failover. Context preserved across providers. |
| Live agent coverage (Brand-trained human support) |
24/7/365 | Median response <3 minutes. 99% of escalations within 5 minutes. |
| Event streaming (Opera Cloud sync) |
99.0% delivery | Queued, idempotent delivery with retry logic. 2xx success within delivery window. |
Oracle Cloud Infrastructure (OCI) and Opera Cloud are third-party services outside CordiantOne's operational control. We implement mitigation strategies (queuing, read-only continuity, live agent escalation) but cannot control upstream restoration windows. Third-party infrastructure outages are excluded from SLA calculations.
Response time commitments
All incidents classified by severity (S1–S4) with committed response and resolution targets. S1/S2 incidents receive live status updates via your preferred channel.
Severity Classification & Response Times
| Severity | Definition | Initial Response | Target Resolution |
|---|---|---|---|
| S1 | Platform down or broad transaction failure. Guest-facing services unavailable or critical commerce functions (booking, payment) completely broken. | 15 minutes (24/7) |
1–4 hours Root Cause Analysis (RCA) provided within 72 hours |
| S2 | Major feature degraded. Significant functionality impaired but workarounds available. Example: Booking flow functional but slow. | 1 hour (24/7) |
4–12 hours |
| S3 | Partial loss with workaround. Minor feature impaired, guest experience slightly degraded but commerce continues. | 4 business hours | 1–3 business days |
| S4 | Minor issue or enhancement request. Cosmetic issues, feature requests, non-urgent improvements. | 1 business day | Next planned release |
All S1 incidents escalate to VP Operations within 30 minutes if initial response doesn't show progress. Executive escalation (CTO) occurs if S1 remains unresolved beyond 2 hours. Customer receives status updates every 30 minutes during active S1 incidents.
When dependencies fail
CordiantOne operates with multiple external dependencies (Opera Cloud, AI providers, OCI). Here's exactly what happens when each fails—and how we maintain service continuity.
What if Opera Cloud is down?
Conversational service continues. CordiantOne remains operational in read-only mode. Guests can ask questions, access information, and browse services.
Transactions pause safely. Deterministic workflows that would write to Opera Cloud (bookings, modifications, payments) pause to preserve data integrity.
Live agents capture requests. When a guest needs a transaction during Opera downtime, our brand-trained agents automatically step in, capture the request, and forward it to your authorized staff via pre-approved channels.
No automatic replay. Once Opera restores, your staff performs manual reconciliation. CordiantOne does not automatically replay paused transactions—safety over convenience.
What if AI providers fail?
Automatic failover. CordiantOne runs multi-provider AI orchestration. If the primary model degrades, we automatically fail over to pre-configured secondary providers.
Context preserved. Conversation history and guest context transfer seamlessly between providers. Guests never notice the switch.
Human continuity. If all configured AI providers become unavailable (rare), CordiantOne transitions to human-assisted mode—brand-trained live agents continue conversations using the same contextual thread until AI restores.
What if OCI is down?
Third-party infrastructure. Oracle Cloud Infrastructure (OCI) is outside CordiantOne's operational control. Regional OCI outages affecting our tenancy will temporarily make CordiantOne services unavailable.
We notify immediately. Customers receive status updates, ongoing progress reports, and coordination assistance during restoration.
SLA exception. OCI outages are treated as third-party infrastructure failures and excluded from CordiantOne uptime calculations (per standard industry practice).
Service credits for SLA breaches
Service credits are the exclusive monetary remedy for validated SLA breaches. Excludes scheduled maintenance, customer misconfiguration, force majeure, and third-party infrastructure outages.
Credit Structure
| SLA Breach Condition | Credit Applied | Notes |
|---|---|---|
| Platform uptime < 99.95% in any calendar month | 10% monthly service credit | Calculated after excluding scheduled maintenance and third-party infrastructure outages. Maximum one credit per breach category per month. |
| AI conversation layer uptime < 99.9% in any calendar month | 10% monthly service credit | Subject to verification via system logs. Multi-provider orchestration failures only (not upstream provider maintenance). |
| Verified live agent coverage lapse (no response within 10 minutes) | 5% monthly service credit | Applied per validated instance after review. Does not apply if guest disconnects or issue self-resolves before agent connects. |
Credit requests must be submitted within 30 days of the incident with supporting logs or evidence. CordiantOne will review and respond within 10 business days. Approved credits applied to following month's invoice. Total credits capped at one month's service fee per product area per calendar quarter.
How to reach us
Multiple channels for different urgency levels. Enterprise customers assigned a Customer Success Account Manager (CSAM) and Solutions Engineer during onboarding.
Can you replay queued transactions after Opera outage?
No. CordiantOne does not automatically replay transactional requests to Opera Cloud after service restoration. Live agents capture requests during outages and forward them to property staff for manual reconciliation—this prioritizes data integrity over automation.
Is multi-LLM failover automatic?
Yes. CordiantOne automatically fails over to configured secondary AI providers using circuit-breaker logic when primary models degrade. Conversation context transfers seamlessly. Customer can specify provider preferences per contract terms.
How are third-party outages treated in SLA calculations?
OCI and Opera Cloud outages are third-party infrastructure exceptions and excluded from CordiantOne uptime calculations. We provide continuous notifications, status updates, and post-incident reports for transparency—but cannot control external infrastructure restoration windows.
What happens during scheduled maintenance?
Scheduled maintenance announced 7 days in advance via email and customer portal. Typical windows: 2am-4am in your property's timezone. Most updates deploy with zero downtime. Scheduled maintenance excluded from SLA availability calculations.