Acceptance and Scope

By accessing the MAD-AI platform ("the Service"), you agree to these Terms. If acting for an organization, you represent you have the authority to bind that entity. The Service provides multi-agent AI automation for 8D, FMEA, and PPAP workflows designed to assist expert human judgment.

Data Ownership and Intellectual Property

Customer Ownership: You retain all rights to uploaded data (ERP logs, sensor telemetry, etc.).

Knowledge Preservation: Ingested "Tribal Knowledge" is treated as Tier-1 Sensitive IP, siloed to your environment, and never used to train models for other customers.

Usage Data: We may use anonymized, de-identified metadata to improve system performance, provided it never identifies the User or proprietary processes.

IATF 16949 Compliance & Reliability

Human-in-the-Loop: AI outputs are advisory. Final IATF 16949 compliance and safety-critical approvals remain the sole responsibility of the User’s qualified personnel.

Audit Readiness: All outputs include a "Digital Paper Trail" documenting data sources and logic.

Record Retention: We adhere to IATF 16949:2016 (Clause 7.5.3.2.1) for secure storage and recovery of quality-critical documentation.

Enterprise Security Protocols

Architecture: We utilize a strict multi-tenant architecture with logical data isolation (no "co-mingling").

Encryption: Data is secured via AES-256 at rest and TLS 1.3 in transit.

Access: We support SSO and MFA to ensure only authorized personnel can access or approve workflows.

Certifications: Our infrastructure undergoes regular SOC2 Type II / ISO 27001 third-party audits.

Liability and Obligations

User Responsibility: Users must ensure MAD-AI usage aligns with internal quality manuals and Customer-Specific Requirements (CSRs).

Limitation of Liability: MAD-AI is not liable for indirect damages, including production downtime or loss of certification. Total liability is capped at the fees paid by the User in the preceding 12 months.