KAIDF-AMM 14-Day Diagnostic Sprint
Comprehensive static analysis and deterministic verification of internal AI codebases. We translate NIST AI RMF, ISO 42001, LGPD, and EU AI Act into active CI/CD guardrails and a Board-Ready Scorecard.
Audit Matrix (5 Core Dimensions)
Automated AST & Static RulesVerification of ingestion boundaries, PII leakage prevention, outbound payload lock, and VPC data residency.
Deterministic pre-commit hooks, runtime fence verification, prompt-injection defenses, and fail-closed CI/CD triggers.
Direct code traceability mapped against NIST AI RMF 1.0, ISO/IEC 42001 controls, EU AI Act High-Risk criteria, and LGPD.
Model dependency pinning, deterministic fallback bounds, local inference capability, and air-gapped readiness.
Enforced token burn ceiling, runaway execution circuits, and edge overhead optimization.
Formal Board-Ready Scorecard (PDF/JSON) + Actionable GitHub engineering remediation backlog.
Sprint Execution Timeline (14 Calendar Days)
Read-only code access configuration (Deploy Key / Tarball) and initial CLI AST baseline test run.
Execution of custom K-AIDF rulesets, dependency vulnerability checks, and unauthorized external call detection.
Generation of the Board-Ready Risk Scorecard and remediation PR backlog ready for engineering sprint ingestion.
60-minute confidential debriefing with CISO, CTO, and Lead Architect Ricardo Filipo for formal handover.
Secured via Stripe Hosted Checkout. Zero tracking scripts or persistent storage.