Contractual 99.95% Service Level Agreement (SLA)
Financially backed 99.95% Monthly Uptime Percentage commitment, P1-P4 engineering escalation matrix, and 15-minute RPO / 2-hour RTO disaster recovery.
Executive Summary & Operational Continuity Guarantee: Publishing houses rely upon British Noble during high-stakes seasonal frontlist embargoes, offset printer slot deadlines, live international book-fair acquisitions, and quarterly or semi-annual financial royalty closes. Accordingly, British Noble backs production workspaces across UK, European, North American, and UAE sovereign vaults under a binding, financially enforced 99.95% Monthly Uptime Percentage commitment (permitting less than twenty-two minutes of unscheduled unavailability in a standard thirty-day month), supported by three-Availability-Zone synchronous replication, zero-downtime blue/green software deployments, a 15-minute Recovery Point Objective (RPO), and a 2-hour Recovery Time Objective (RTO).
Global incident response is staffed directly by publishing systems engineers fluent in EDItEUR ONIX 3.1.2 XML schemas, 42-rule PDF/X-1a prepress pipelines, Adobe IDML geometry, and multi-currency royalty ledgers. Severity 1 critical production incidents trigger a ≤ 15-minute initial engineering response (24/7/365) for Enterprise customers with 30-minute status updates, backed by automatic telemetry attribution, tiered monthly fee credits of 10%, 20%, or 30%, and an explicit penalty-free termination and refund right should chronic availability degradation ever occur. Routine software releases deploy without planned weekday interruptions, mandatory book-fair change freezes safeguard peak seasonal publishing windows, and blameless Post-Incident Reviews deliver root-cause engineering transparency within five business days via Support@BritishNoble.com.
1. Scope of Contractual 99.95% Availability Commitment
This Contractual Service Level Agreement ("SLA") forms a binding component of the Master SaaS & Author Studio Terms of Service for all paid production subscriptions of British Noble AI and active Connected Author Portals. British Noble guarantees that Production Workspaces within Customer's designated sovereign cloud vault (UK, EU, US, or UAE) shall achieve a Monthly Uptime Percentage of at least 99.95% during every calendar month of the Subscription Term.
"Core Production Services" covered by this availability commitment comprise: (a) authenticated workspace sign-in and Single Title Record relational read/write operations; (b) Editorial Intelligence document access and version sign-off gates; (c) Book Production Studio layout rendering and 42-rule preflight verification queues; (d) Distribution Network ONIX 3.1.2 XML compilation and outbound SFTP/API dispatch queues; (e) Sales & Royalties calculation and statement generation engines; and (f) Connected Author Portal proof review and statement access.
2. Monthly Uptime Percentage Mathematical Formula & Probes
Monthly Uptime Percentage is measured independently per sovereign cloud region and per tenant cluster across each calendar month in Coordinated Universal Time (UTC) using the following deterministic formula:
Monthly Uptime Percentage = ((Total Calendar Minutes in Month - Unscheduled Downtime Minutes) / Total Calendar Minutes in Month) × 100
"Unscheduled Downtime" means any period of two (2) or more consecutive minutes during which synthetic multi-region health probes (executing every 60 seconds across authentication, relational database commit, object storage retrieval, and OpenAPI 3.1 endpoints) or verified tenant telemetry record an HTTP 5xx server failure rate exceeding 2.0% of valid requests, or complete inability for authorised users to authenticate and access Single Title Records, excluding solely the narrowly defined Exclusions in Clause 10. Partial degradation affecting solely an isolated module (such as ONIX batch dispatch while Editorial and Royalty modules remain 100% operational) is weighted conservatively at fifty percent (50%) of elapsed incident minutes toward Unscheduled Downtime so module-level regressions remain financially accountable.
3. Tiered Financial Service Credits & Chronic Breach Termination
If the Monthly Uptime Percentage for Customer's production workspace falls below 99.95% in any calendar month, Customer is entitled to receive a financial Service Credit calculated as a percentage of the pro-rated monthly subscription fee attributable to the affected workspace for that calendar month (one-twelfth of the annual subscription fee):
| Achieved Monthly Uptime Percentage | Approximate Monthly Unscheduled Window | Contractual Service Credit |
|---|---|---|
| 99.50% to < 99.95% | > 21.9 minutes up to 3.6 hours | 10% of Monthly Subscription Fee |
| 99.00% to < 99.50% | > 3.6 hours up to 7.3 hours | 20% of Monthly Subscription Fee |
| Below 99.00% | Exceeding 7.3 hours in a calendar month | 30% of Monthly Subscription Fee |
Chronic Availability Failure Exit Right: Enterprise publishing teams cannot tolerate recurring operational instability even with invoice credits. Accordingly, if Monthly Uptime Percentage falls below 99.00% in three (3) consecutive calendar months, or in any four (4) calendar months within a rolling twelve-month period, Customer may terminate the Agreement for material cause upon thirty (30) days' written notice, trigger immediate zero-fee full catalogue export, and receive a full cash refund of all prepaid, unearned subscription fees for the remainder of the Subscription Term.
4. Automated Telemetry Verification & Credit Application
Whenever British Noble's observability platform records an SLA breach affecting a production tenant, our Site Reliability Desk logs an automatic SLA Incident Ticket and notifies Customer's administrative contact. To formalise a Service Credit request, Customer submits a brief written notice citing the calendar month and Incident ID to Support@BritishNoble.com within thirty (30) calendar days following the close of the affected calendar month.
British Noble verifies probe logs within five (5) business days and applies approved Service Credits directly against Customer's next renewal or expansion invoice. Where an approved Service Credit arises during the final month of a non-renewing Subscription Term, British Noble remits the equivalent amount as a direct monetary refund to Customer's bank account within thirty (30) calendar days.
5. Incident Severity Classification & Initial Response SLAs
Support enquiries and operational incidents are classified into four objective severity tiers. Enterprise & Multi-Imprint Group editions include 24/7/365 P1 critical paging and named solutions pods, while Mid-Market and Growth editions receive accelerated regional coverage:
| Severity Level | Operational Definition | Enterprise Initial SLA | Mid-Market / Growth | Update Cadence |
|---|---|---|---|---|
| Severity 1 (P1 - Critical) | Complete workspace outage, data-integrity risk, security incident, or live embargoed ONIX / royalty payout batch blocked on deadline | ≤ 15 Minutes (24/7/365) | ≤ 1 Hour (Mid-Market) ≤ 4 Hours (Growth) | Every 30 minutes until mitigated |
| Severity 2 (P2 - High) | Major capability degraded (e.g., PDF/X-1a preflight worker stall or ERP sync failure) with no immediate operational workaround | ≤ 1 Hour (24/7) | ≤ 3 Hours (Mid-Market) ≤ 6 Hours (Growth) | Every 2 hours until resolved |
| Severity 3 (P3 - Medium) | Non-critical feature defect, single-connector warning, or UI anomaly where an operational workaround is available | ≤ 4 Business Hours | ≤ 8 Business Hours | Once per business day |
| Severity 4 (P4 - Advisory) | ONIX 3.1.2 codelist mapping guidance, trim/bleed template questions, sandbox setup, or feature enhancement request | ≤ 1 Business Day | ≤ 1 Business Day | As mutually scheduled |
Customer administrators may upgrade the severity classification of any open ticket at any time if an approaching publication embargo, book-fair deadline, or financial period close increases business impact.
6. Engineering Escalation Path & Book-Fair Change Freezes
Unlike generic outsourced call centres, Tier 1 intake at British Noble is staffed directly by publishing systems engineers trained in XML schemas, Adobe IDML geometry, ICC colour management, and double-entry royalty accounting. If any P1 incident is not mitigated within thirty (30) minutes of opening, automated PagerDuty escalation engages the on-call Duty Incident Commander; at sixty (60) minutes, the relevant Principal Domain Architect (Prepress, ONIX Distribution, or Financial Ledgers) joins the dedicated incident bridge; and at ninety (90) minutes, the VP of Infrastructure and CISO assume executive oversight.
Seasonal Publishing Change Freezes: Because global trade and scholarly publishing concentrates mission-critical dispatches around international fairs and Q4 holiday peaks, British Noble enforces mandatory non-emergency infrastructure change freezes during the week preceding and throughout the London Book Fair, Frankfurt Book Fair, Sharjah International Book Fair, and the mid-November to early-December peak retail fulfilment window.
7. Multi-AZ High Availability & Disaster Recovery Commitments
Every regional vault (eu-west-2 London, eu-central-1 Frankfurt, us-east-1 Virginia, and me-central-1 Dubai) runs active-active stateless application clusters across a minimum of three physically independent Availability Zones separated by distinct power grids, flood plains, and fibre paths, backed by synchronous Multi-AZ relational database clusters and eleven-nines durability encrypted object storage.
In the event of a catastrophic single-Availability-Zone failure, automated load-balancer and database failover completes transparently within ninety (90) seconds with zero data loss. For catastrophic regional disaster scenarios, British Noble maintains a contractually binding 15-minute Recovery Point Objective (RPO) via continuous write-ahead log streaming and encrypted point-in-time snapshots, paired with a 2-hour Recovery Time Objective (RTO). Full end-to-end disaster-recovery and failover drills are executed bi-annually (every May and October), with sanitised RPO/RTO timing verification reports available to Enterprise customers upon request.
8. Zero-Downtime Releases & Scheduled Maintenance Rules
Routine application deployments, preflight rule updates, and EDItEUR codelist refreshes utilise blue/green rolling container orchestration and backward-compatible database migrations, resulting in zero planned downtime during the working week across UK, European, Americas, and MENA business hours.
On the rare occasion that major underlying database engine or HSM firmware maintenance requires brief write-quiescence, such Scheduled Maintenance: (a) shall occur exclusively between Saturday 22:00 UTC and Sunday 04:00 UTC; (b) shall be announced to workspace administrators via email and in-app banner at least seven (7) calendar days in advance (unless an emergency zero-day cryptographic vulnerability mandates urgent patching, in which case maximum practicable notice is given); and (c) shall never exceed four (4) cumulative hours in any calendar quarter. Any planned maintenance overrunning its announced window is counted strictly as Unscheduled Downtime.
9. Asynchronous Preflight, ONIX & Royalty Compute Targets
Beyond core HTTP availability, publishing velocity depends upon predictable background job execution. Under normal operational load within edition concurrency limits, British Noble engineers asynchronous worker auto-scaling to meet the following p95 processing targets:
- 42-Rule Print & Digital Preflight QC: Completion within five (5) minutes for standard monochrome/trade PDF/X-1a and EPUB 3.3 files, and within twelve (12) minutes for high-resolution 254 x 305 millimetre illustrated coffee-table monographs exceeding 1.5 GB.
- ONIX 3.1.2 Delta Compilation & Queue Dispatch: Schema validation and outbound partner SFTP/API transmission initiated within three (3) minutes of human release approval.
- Royalty Period Calculation Runs: Full multi-tier escalator, reserve, and FX computation completed within ten (10) minutes for catalogues up to 10,000 ISBNs and five million sales lines.
10. Narrowly Tailored SLA Exclusions
Unscheduled Downtime calculations exclude unavailability solely and directly caused by: (a) Scheduled Maintenance conducted strictly in compliance with Clause 8; (b) Force Majeure events outside British Noble's reasonable infrastructure redundancy control (such as submarine cable cuts affecting external ISP transit outside cloud backbone peering); (c) outages or misconfigurations in Customer's own corporate network, workstation browsers, or Customer-administered SAML Identity Provider (Okta, Entra ID, Google Workspace); (d) external third-party retailer or wholesaler SFTP server downtime (e.g., where British Noble's ONIX queue retries cleanly with exponential backoff while an external retail partner's intake server is offline); (e) 14-day free evaluation sandboxes; or (f) temporary emergency suspension of a specific API key generating abusive runaway loops in violation of documented rate limits.
11. Post-Incident Reviews (PIRs) & Operational Governance
Following any Severity 1 or Severity 2 production incident, British Noble's Site Reliability Desk delivers a preliminary incident summary within twenty-four (24) hours and a comprehensive written Blameless Post-Incident Review (PIR) within five (5) business days. Every PIR documents an exact minute-by-minute timeline, root-cause architectural analysis, customer impact telemetry, detection latency, and tracked engineering preventative actions with target completion dates.
Enterprise & Groups customers receive formal Quarterly Operational Reviews led by their named Technical Account Manager, reviewing achieved regional uptime (historically ≥ 99.98%), ONIX partner acceptance rates across 108 channels, preflight pass ratios, and upcoming roadmap readiness.
Multi-Jurisdiction Governance
Aligned with UK GDPR, Data Protection Act 2018, EU SCCs, CCPA/CPRA, UAE Federal Decree-Law No. 45 of 2021 (PDPL), ISO/IEC 27001:2022, and SOC 2 Type II.
Zero Manuscript Training
Customer manuscripts, proof files, ONIX records, and royalty ledgers are isolated per tenant and never used to train foundation or shared models.
Site Reliability & Publisher Support Desk
Enterprise procurement teams may request countersigned schedules, DPA exhibits, and audit certificates directly.
Support@BritishNoble.com →