SLA & Uptime Policy

k&z uptime commitments, measurement methodology, and service credit provisions for our quantum AI infrastructure platform.

Effective Date: January 1, 2026 — Version: 1.0

1. Introduction

This Service Level Agreement and Uptime Policy (“SLA”) describes the uptime commitments, performance standards, and service credit provisions that K&z Limited (“k&z”) provides to customers of the k&z quantum AI infrastructure platform. This SLA is incorporated into and forms part of the Terms of Service and any applicable Master Service Agreement (“Agreement”).

k&z recognizes that reliability and availability are critical to our clients’ research and operations. We are committed to maintaining high service availability and providing transparent measurement and remediation when service levels are not met.

2. Definitions

  • “Downtime” means the total number of minutes during a calendar month in which the applicable Service is unavailable, as measured by k&z’s monitoring systems. A Service is considered unavailable when a client is unable to submit jobs to or retrieve results from the QPU resources or API endpoints designated for their use, excluding Scheduled Maintenance and Exclusions defined herein.
  • “Monthly Uptime Percentage” means the total number of minutes in a calendar month minus Downtime, divided by the total number of minutes in the calendar month, expressed as a percentage.
  • “Scheduled Maintenance” means planned maintenance activities for which k&z provides at least seventy-two (72) hours’ prior notice through the k&z status page and email notification.
  • “Service Credit” means a credit against future fees, calculated as a percentage of the monthly fees for the affected Service.
  • “Error Rate” means the percentage of API requests that result in server errors (HTTP 5xx responses) during a given measurement period.

3. Uptime Commitments

3.1 Service Tiers

k&z provides the following uptime commitments based on the service tier:

Service Component Service Tier Monthly Uptime Commitment
Quantum Runtime API Standard 99.95%
Quantum Runtime API Enterprise 99.99%
Shared QPU Resources Standard 99.90%
Dedicated Single-Tenant QPU Enterprise 99.95%
Platform Dashboard & Management Console All Tiers 99.95%
Hybrid Quantum + HPC Orchestration Standard 99.90%
Hybrid Quantum + HPC Orchestration Enterprise 99.95%
Sovereign Deployments Sovereign Per contract (minimum 99.95%)

3.2 API Performance

In addition to uptime commitments, k&z targets the following API performance standards:

  • API Latency: Median API response time of less than 200 milliseconds for job submission and status queries; 95th percentile response time of less than 500 milliseconds.
  • Error Rate: Monthly average API error rate (5xx responses) of less than 0.1% for Standard tier and less than 0.01% for Enterprise tier.
  • Job Queue Time: For on-demand QPU access, target queue wait time of less than 60 seconds for Standard priority and less than 10 seconds for Priority access.

4. Measurement Methodology

4.1 Monitoring

k&z measures service availability using a combination of internal monitoring systems, including:

  • Synthetic Monitoring: Automated health checks performed every sixty (60) seconds from multiple geographic locations, simulating client API requests and job submissions.
  • Real User Monitoring: Aggregated analysis of actual client API requests and their success/failure rates.
  • Infrastructure Monitoring: Continuous monitoring of QPU hardware status, cryogenic system health, control electronics, network connectivity, and compute cluster availability.

4.2 Downtime Calculation

Downtime is measured from the moment k&z’s monitoring systems detect that the Service is unavailable until the moment monitoring confirms restoration of normal operation. Partial degradation that does not prevent clients from submitting jobs or retrieving results is not counted as Downtime, though it may trigger performance-related remediation under Section 3.2.

4.3 Monthly Uptime Percentage Formula

Monthly Uptime Percentage is calculated as follows:

Monthly Uptime % = ((Total Minutes in Month − Downtime Minutes) ÷ Total Minutes in Month) × 100

For example, in a month with 43,200 total minutes (30 days) and 20 minutes of Downtime, the Monthly Uptime Percentage would be: ((43,200 − 20) ÷ 43,200) × 100 = 99.954%.

5. Service Credits

5.1 Credit Schedule

If k&z fails to meet the applicable Monthly Uptime Commitment, clients are eligible for Service Credits as follows:

Monthly Uptime Percentage Service Credit (% of Monthly Fees)
99.00% to < Uptime Commitment 10%
95.00% to < 99.00% 25%
90.00% to < 95.00% 50%
Below 90.00% 100%

5.2 Calculation Basis

Service Credits are calculated as a percentage of the monthly fees actually paid by the client for the specific Service component affected by the Downtime during the month in which the Downtime occurred. For consumption-based billing, the credit is calculated based on the average monthly fees over the preceding three (3) months, or the actual fees for the affected month, whichever is greater.

5.3 Application of Credits

Service Credits are applied to the client’s next invoice. Credits are not transferable, may not be exchanged for cash, and expire twelve (12) months after issuance if not applied.

6. Exclusions

The following events and circumstances are excluded from Downtime calculations and do not give rise to Service Credits:

  • Scheduled Maintenance: Planned maintenance for which k&z provides at least seventy-two (72) hours’ prior notice. k&z schedules routine maintenance during low-usage periods and targets no more than four (4) hours of scheduled maintenance per month.
  • Force Majeure Events: Events beyond k&z’s reasonable control, including natural disasters, war, terrorism, government actions, pandemics, power grid failures, and internet backbone failures.
  • Client-Caused Issues: Downtime caused by the client’s actions, including misconfiguration, excessive API calls that trigger rate limiting, violations of the Acceptable Use Policy, or failure to follow published integration guidelines.
  • Third-Party Failures: Failures in third-party systems, networks, or software not under k&z’s direct control, including internet service provider failures and DNS resolution issues.
  • Quantum Hardware Events: Downtime resulting from inherent quantum hardware phenomena, including unexpected qubit decoherence events, cryogenic system failures requiring extended cool-down periods, and calibration cycles necessitated by environmental changes, provided k&z demonstrates that reasonable maintenance and monitoring practices were followed.
  • Beta Services: Any services designated as beta, preview, or early access are excluded from SLA commitments.
  • Suspended Accounts: Periods during which the client’s account is suspended for non-payment or policy violations.

7. Claim Process

7.1 Filing a Claim

To receive a Service Credit, the client must submit a claim to k&z within thirty (30) days of the end of the calendar month in which the Downtime occurred. Claims must be submitted via email to support@kandz.co or through the platform support portal and must include:

  • The client’s account identifier and organization name;
  • The date(s) and time(s) of the claimed Downtime;
  • The specific Service component(s) affected;
  • A description of the impact experienced; and
  • Any supporting evidence, such as error logs, screenshots, or API response records.

7.2 Review and Response

k&z will review the claim against its internal monitoring data and respond within fifteen (15) business days. If the claim is validated, the Service Credit will be applied to the next invoice. If the claim is denied, k&z will provide a written explanation including relevant monitoring data.

7.3 Dispute Resolution

If the client disagrees with k&z’s determination, the client may escalate the dispute by providing additional evidence within fifteen (15) business days of receiving k&z’s response. k&z will conduct a secondary review and provide a final determination within fifteen (15) additional business days. Further disputes shall be resolved in accordance with the dispute resolution procedures in the applicable Agreement.

8. Credit Limits

8.1 Maximum Credits

The maximum aggregate Service Credits issued to a client in any calendar month shall not exceed one hundred percent (100%) of the monthly fees paid by the client for the affected Service component(s) during that month.

8.2 Sole Remedy

Service Credits are the client’s sole and exclusive remedy for k&z’s failure to meet the service level commitments set forth in this SLA. Service Credits do not limit or waive any other rights the client may have under the Agreement, including termination rights for persistent service failures.

8.3 Persistent Failures

If k&z fails to meet the applicable Monthly Uptime Commitment for three (3) or more consecutive months, the client may terminate the affected Service(s) upon thirty (30) days’ written notice and receive a pro-rata refund of any prepaid fees for the unused portion of the committed term.

9. Status Page and Notifications

9.1 Service Status

k&z maintains a public service status page at status.kandz.co that provides real-time information about the operational status of all Service components. Clients are encouraged to subscribe to status updates via email, RSS, or webhook.

9.2 Incident Communication

During service disruptions, k&z will post updates to the status page at least every thirty (30) minutes until the incident is resolved. Post-incident reports for significant outages will be published within five (5) business days of resolution.

10. Contact

For SLA-related inquiries and claims: