📞 (84) 0968 695 865 · contact@dbiz.com
Tuyển dụng Tài nguyên VI | EN
Demo Dùng thử miễn phí

Security & compliance — trust is the foundation of the AI-native enterprise

Security & compliance

Trust is the foundation of the AI-native enterprise

Your data and your models are protected to international standards, with AI operating under control and people always holding the decision.

ISO/IEC 27001 Decree 13/2023 Human-in-the-loop Data sovereignty
Data sovereignty

Data and AI models reside within your enterprise — under your full control, with no dependency on a third party.

Encryption & permissions

Encrypted at rest and in transit; granular row- and column-level permissions (RLS/CLS) and role-based access.

Logging & traceability

Every user action and every AI agent action is logged, traceable and auditable.

Human-in-the-loop

AI proposes — a person approves at every point of risk. With stop points, recovery and cost limits.

AI guardrails

Constraints on the scope of action, output screening, and prevention of sensitive data leakage.

Backup & recovery

Automated backup and a business continuity and disaster recovery plan (BCP/DR) to keep operations running.

Compliance with Decree 13/2023

Meeting Vietnam’s personal data protection requirements across the entire data lifecycle.

ISO/IEC 27001

An information security management system (ISMS) to the international standard.

A layered AI-native architecture

Each layer is isolated, secured and scaled independently — keeping data, models and agents fully controlled and traceable.

FoundationCore platform & identity
Data LakehouseUnified data, permissioned
Agent HarnessControlled AI with human-in-the-loop
InfrastructureCPU · GPU · private cloud
Request the security whitepaper →

Platform products meeting the technical requirements of ISO/IEC 27001 & 42001

Every DBIZ platform product is designed with technical controls mapped to ISO/IEC 27001 (information security) and ISO/IEC 42001 (AI governance). The table below summarises how each component meets them.

Platform productMeeting ISO/IEC 27001 — information securityMeeting ISO/IEC 42001 — AI governance
DBIZ Foundation
Core platform & identity
  • Centralised identity and access management (SSO, RBAC, MFA)
  • Multi-tenant isolation and environment separation
  • Key and secrets management
  • Centralised audit logging across the system
  • A secure software development lifecycle (SSDLC)
  • A model and agent registry with named owners and an approval process
  • A central AI policy enforcement point
  • Defined roles and accountability
DBIZ Autonomous
Controlled AI agents with human-in-the-loop
  • Least-privilege permissions for each agent
  • Guardrails constraining the scope of action
  • Output screening and prevention of sensitive data leakage
  • Logging and traceability of every agent action
  • A secrets vault for the tools an agent calls
  • Human-in-the-loop approval at every point of risk
  • AI risk and impact assessment
  • Stop points, recovery (kill-switch/rollback) and cost limits
  • Decision logging and explainability
  • Monitoring of behavioural and model drift
DBIZ Data Lakehouse
Unified data, permissioned
  • Encryption at rest and in transit
  • Granular row- and column-level permissions (RLS/CLS)
  • Data classification, masking and anonymisation
  • Full lineage and access logging
  • Automated backup and disaster recovery (DR)
  • Data governance and quality for AI
  • Lineage of training data
  • Minimisation of personal data (PII) used for AI
  • Dataset documentation supporting bias control
DBIZ Blockchain
Immutability & provenance
  • An immutable, tamper-evident log
  • Cryptographic integrity and non-repudiation
  • Permissions and controls enforced in smart contracts
  • Attestation of the provenance of AI decisions (model and data attestation)
  • An immutable ledger for end-to-end traceability and audit of AI actions
AI modules & 97 AI agents
AI business modules
  • Inheriting Foundation controls (identity, permissions, encryption, audit)
  • Data scope and access rights limited per module
  • Each module has a documented purpose, data scope and monitoring mechanism
  • People control and monitor output quality, with alerts on anomalies

Shared foundation: CPU, GPU and private cloud infrastructure assuring data and model residency — the basis for both ISO/IEC 27001 and 42001.