# Qurmmius Intelligence Layer — customer quickstart

This page is for an authorized Qurmmius Intelligence workspace. It explains the production path without exposing credentials or provider payloads.

## What is available

- **Starter — $49/month:** one product skill.
- **Studio — $149/month:** three product skills.
- **Platform — $399/month:** up to ten skills.

QI routes redacted requests through a protected server gateway when provider mode is enabled. Local fallback remains available when a provider is unavailable, and raw prompts, media, credentials, and personal data are not written to browser history.

## First-use checklist

1. Open [QI Intelligence](/intelligence/) and sign in to the authorized workspace.
2. Choose the product skill and submit a synthetic or appropriately redacted request first.
3. Review the structured result, confidence, privacy mode, request ID, and recommended next action.
4. For a paid plan, review the [QI Terms](/qi-terms.html) and [Refund & Cancellation Policy](/refund-policy.html) before intentionally starting checkout.
5. After payment, use the Stripe-hosted customer portal from the account controls to update billing details or cancel a renewal.

## Support and incident reporting

Contact [Qurmmius support](/#contact) or email [qurmmie@gmail.com](mailto:qurmmie@gmail.com). Include the workspace name, approximate time, request ID, product skill, and a short description. Do **not** send API keys, card numbers, access codes, raw personal data, or unredacted media.

For a suspected billing issue, include the invoice or Stripe receipt number. Billing requests are acknowledged within three business days and normally reviewed within ten business days under the published refund policy.

For a suspected service incident, check the non-secret [QI health endpoint](/api/qi-health) and attach the returned status, timestamp, and request ID. Health output never contains secrets.

## Operational boundaries

QI outputs are decision support, not a guarantee of a racing result, grant result, image reconstruction, safety outcome, or business outcome. Human review remains required for consequential decisions. DeepFake Shield stays purpose-limited and local-preview only, and LegacyVault remains synthetic dry-run only.

## Public references

- [Production readiness gate](/QI-PRODUCTION-READINESS.md)
- [Billing activation checklist](/QI-BILLING-ACTIVATION.md)
- [Privacy & data-handling notice](/privacy/)
- [QI Terms](/qi-terms.html)
- [Refund & Cancellation Policy](/refund-policy.html)