AI System Disclosure & Responsible-Use Statement

1. What Quovec Is — and Is Not

Quovec™ is a decision-support platform for multi-unit business operators. It analyzes the customer's own operational data (sales, labor, inventory, leases) and surfaces findings, scores, and recommendations for human review.

• It is not an autonomous decision-maker. No payment, price change, personnel action, schedule change, or external communication is executed by the system without approval by an authorized human at the customer.
• It is not an employee-surveillance or lie-detection tool, and it makes no determination about any individual's honesty, intent, or conduct.
• It is not a substitute for professional advisors (legal, tax, accounting, HR).

2. How the Intelligence Works (Deterministic by Design)

Confidence floors. The system declines to speak rather than guess: findings with insufficient data return "needs data" instead of a low-quality answer; forecast inputs carry confidence values, and — as a matter of platform policy — weather forecasts below 85% confidence are excluded from staffing recommendations entirely.

No training on customer data. Model providers processing data through the Service are contractually restricted from using it to train their models.

3. Employee-Related Analytics — Special Disclosure

Some features surface statistical variances that may be associated with locations, shifts, or personnel (for example, loss-prevention or attendance patterns). Customers must understand and agree:

• Findings are signals for review — never accusations. A variance is a statistical observation with documented method and confidence, not a determination of misconduct.
• Cohort-controlled fairness. Person-associated comparisons are made against matched peer cohorts (e.g., same role/shift context), not raw averages, to reduce unfair attribution.
• Human decision required. The platform makes no automated employment decision and is not designed, marketed, or permitted to be used as an automated employment decision tool. Any investigation, discipline, or termination is the customer's decision, made under the customer's own policies and applicable employment law, with independent verification.
• Customer responsibilities. The customer is responsible for providing any notices to or obtaining any consents from its workforce required by applicable law, and for the lawful use of outputs.

4. Accuracy, Limits & Appropriate Reliance

• Output quality depends on the accuracy, completeness, and freshness of the customer's data. Broken feeds, missed entries, or unrecorded events degrade results; the system flags data-coverage gaps where it can detect them.
• Forecasts are probabilistic estimates of an uncertain future; historical patterns may not persist. Savings projections are modeled estimates, not guarantees.
• Recommendations should be validated against the customer's own knowledge of its business before acting; the platform exposes its basis (inputs, method, confidence) for every finding so results can be checked.

5. Security & Privacy Posture (Summary)

Per-tenant isolation enforced at the database (row-level security on every tenant table, verified by automated tests); region-scoped access for managers; asymmetric-key authentication; deny-by-default action gating with human approval on sensitive actions; encryption in transit and at rest; tamper-evident audit logging; per-tenant resource budgets. Details: Quovec DPA (Annex A) and the Architecture & Security whitepaper.

6. Human Oversight Commitments

7. Questions & Contact

Technical or compliance questions about this disclosure: security@quovec.com. This disclosure is reviewed at least annually and upon material system changes; the current version is always available at quovec.com/ai-disclosure.