Service Level Agreement

This Service Level Agreement ("SLA") is part of the Terms of Service (the "Agreement") between you and Prediction Lab, LLC ("Prediction Lab," "we," "us," or "our") and describes our uptime commitments and service credit policies for the Cloud Services.

This SLA applies only to paid Cloud Services subscriptions (Multi-Tenant and Single-Tenant deployments). It does not apply to free tiers, trials, BYOC deployments, or the Desktop Application.

1. Definitions

"Downtime" means a period during which the Cloud Services are materially unavailable to you, as determined by our monitoring systems, measured in minutes.

"Monthly Uptime Percentage" means (Total Minutes - Downtime Minutes) / Total Minutes × 100 for a calendar month.

"Service Credit" means a credit calculated as a percentage of your monthly fees for the affected Cloud Services.

"Scheduled Maintenance" means maintenance for which we provide at least 48 hours advance notice.

2. Uptime Commitment

Tier	Commitment	Approximate Allowed Downtime/Month
Standard	99.5%	3.6 hours
Enterprise	99.9%	43 minutes

3. Service Credits

Standard Tier (99.5% Commitment)

Monthly Uptime	Service Credit
99.0% to < 99.5%	10% of monthly fees
95.0% to < 99.0%	25% of monthly fees
< 95.0%	50% of monthly fees

Enterprise Tier (99.9% Commitment)

Monthly Uptime	Service Credit
99.5% to < 99.9%	10% of monthly fees
99.0% to < 99.5%	25% of monthly fees
95.0% to < 99.0%	50% of monthly fees
< 95.0%	100% of monthly fees

4. Credit Limitations

4.1 Monthly Cap. Maximum Service Credit per month is 100% of monthly fees for affected Services.

4.2 No Cash. Service Credits apply to future invoices only, are not redeemable for cash, and are non-transferable.

4.3 Expiration. Credits must be used within 12 months and expire upon subscription termination.

5. Exclusions

The following are not counted as Downtime:

5.1 Scheduled Maintenance. Maintenance with at least 48 hours notice, limited to 4 hours per month, typically during low-usage periods.

5.2 Emergency Maintenance. Critical security patches with as much notice as reasonably practicable.

5.3 External Factors. Internet connectivity issues between your network and our data centers, third-party service failures (except our Subprocessors), DNS issues outside our infrastructure, DDoS attacks, and force majeure events.

5.4 Customer-Caused Issues. Downtime caused by your actions, configurations, equipment, exceeding usage limits, or violations of the Agreement or Acceptable Use Policy.

5.5 AWS Infrastructure. Downtime caused by AWS regional outages affecting underlying infrastructure.

5.6 Excluded Services. Free tiers, trials, BYOC deployments, Desktop Application, Beta features, and deprecated APIs.

6. Requesting Service Credits

6.1 Incident Reporting. Report Downtime by opening a support ticket within 7 days of the incident.

6.2 Credit Request. Submit a written request to support@predictionlab.ai within 30 days after the month in which Downtime occurred, including: account name and contact information, dates and times of incidents, description of impact, and relevant support ticket numbers.

6.3 Verification. We verify Downtime using our monitoring systems. We will work in good faith to resolve discrepancies.

6.4 Processing. Valid requests are processed within 30 days. Approved credits apply to your next invoice.

7. Sole Remedy

SERVICE CREDITS ARE YOUR SOLE AND EXCLUSIVE REMEDY FOR FAILURE TO MEET THE UPTIME COMMITMENT. This SLA does not modify other provisions of the Agreement, including limitations of liability.

8. Termination for Persistent Failures

If Cloud Services fail to meet the applicable Uptime Commitment for 2 out of any 3 consecutive months, and you reported the failures within 7 days of each month's end, you may terminate the affected subscription by providing written notice within 30 days after the third month. You will receive a pro-rata refund of prepaid fees.

9. Service Status

9.1 Status Page. Current status and historical data available at status.predictionlab.ai.

9.2 Incident Notifications. Via status page, email to administrators, and in-product notifications.

9.3 Post-Incident Reports. For significant incidents affecting multiple customers, we provide reports within 5 business days covering timeline, root cause, remediation, and preventive measures.

10. Support Response Times

Standard Tier

Severity	Definition	Response
Critical	Service unavailable, no workaround	4 business hours
High	Major feature impaired	8 business hours
Normal	Minor issue, workaround available	24 business hours
Low	Question or enhancement request	48 business hours

Enterprise Tier

Severity	Definition	Response	Availability
Critical	Service unavailable, no workaround	1 hour	24/7
High	Major feature impaired	4 business hours	Extended hours
Normal	Minor issue, workaround available	8 business hours	Business hours
Low	Question or enhancement request	24 business hours	Business hours

Business Hours: Monday–Friday, 9 AM–5 PM Central Time, excluding US federal holidays.

Extended Hours: Monday–Friday, 7 AM–9 PM Central Time.

Response times are targets for initial acknowledgment, not resolution guarantees.

11. Data Durability

11.1 Backup Frequency. Customer Data backed up at least every 24 hours.

11.2 Backup Retention. Backups retained for at least 30 days.

11.3 RPO. Recovery Point Objective: no more than 24 hours prior to incident.

11.4 RTO. Recovery Time Objective: service restoration within 4 hours.

11.5 Disclaimer. We implement industry-standard procedures but do not guarantee against data loss. Maintain your own backups using our export tools.

12. BYOC Deployments

BYOC deployments are not covered by this SLA. Availability depends on your AWS configuration and operational practices. See our Shared Responsibility Model for details.

13. Changes

We may modify this SLA. Changes that materially reduce commitments will not apply to your current Subscription Term. We provide at least 30 days notice of material changes.

14. Contact

Support: support@predictionlab.ai

Status Page: status.predictionlab.ai