Data Modeling and Data Governance.

Two integrated modules, both supported by AI. Explore model engineering and data management, from inventory to privacy.

MODULE 01

Data Modeling

From visual editing to documentation, with contextual AI and organization-specific standards.

01 / VISUAL EDITOR

Diagramming that handles the real thing.

Modern ER editor built for large models, with version governance and efficient search.

Crow's Foot notation

International standard for ER diagrams. Cardinality and optionality represented unambiguously.

Hierarchical modules

Group objects by domain (HR, finance, etc.) with navigation independent of the macro diagram.

Global search

Find any object or attribute in milliseconds. Works in models with over a thousand tables.

Integrated versioning

Change history with visual diff between versions. Commit comments. Granular rollback.

Scale performance

Diagrams with hundreds of entities without slowdown. Incremental rendering as you zoom.

Automatic layout

Diagram organization algorithms. Suggests arrangements that reduce line crossings.

Honest considerations

The editor is desktop-first. It works on tablets with productivity limitations. Mobile modeling is not an intended use case.

Real-time collaborative editing (multiple users on the same diagram simultaneously) is on the roadmap, not yet available.

02 / MULTI-DATABASE DDL

One model, 13 targets.

Configurable DDL generation for major relational and analytical databases on the market.

Supported databases

Oracle, PostgreSQL, SQL Server, MySQL, MariaDB, DB2, Snowflake, BigQuery, Redshift, Synapse, Teradata, SQLite, MongoDB.

Naming standards

Configure prefixes, suffixes, and conventions per organization. Automatic application during generation.

Incremental scripts

ALTER statement generation based on version diffs. No need to rebuild everything from scratch.

Per-database configuration

Tablespaces, partitioning, specific indexes, and canonical types can be configured per target.

Canonical types

Define abstract types (e.g., "monetary_value") that translate to the native type of each database.

Customizable templates

Generation templates can be adjusted to reflect internal standards without rewriting the tool.

Honest considerations

Legacy or niche databases (e.g., Informix, Sybase ASE) may require adaptation. Evaluated case by case.

Extreme proprietary features (e.g., database-specific geospatial types) may not have direct equivalents in other targets — generation explicitly warns.

03 / ENTERPRISE AUTH

Real SSO. Not a wrapper.

Three native enterprise protocols, with role-based access control and full audit.

OIDC

OpenID Connect with Keycloak, Auth0, ADFS, Okta, and other compatible providers. Claims mapping.

LDAP / Active Directory

Direct corporate directory connection. Customizable filters. LDAPS support.

SAML 2.0

SAML federation with attribute-to-role mapping. Multiple IdPs supported.

Granular RBAC

Customizable roles per organization. Permissions by module, object, and operation.

Automatic provisioning

First-time authenticated users are provisioned with a configurable default role.

Audit trail

Login, model changes, DDL generation, exports — all recorded with timestamp and user.

Honest considerations

MFA (multi-factor) is the corporate IdP's responsibility — 4dbAI delegates authentication to the configured provider.

Direct Kerberos (without federation via LDAP/SAML) is under analysis for the roadmap.

04 / INTEGRATED AI

An assistant that understands your model.

AI grounded in the modeling task — not a generic chat sitting next to it.

Context-aware analysis

The AI "sees" the current model and responds based on it. Not a generic documentation search.

Standard validation

Compares naming, types, and structures against the organization's configured standard.

Improvement suggestions

Identifies missing relationships, recommended normalization, useful indexes.

Natural language

Commands like "add FK from ORDER to CUSTOMER" are executed as real actions on the canvas.

Flexible provider

Connects with your corporate LLM or runs local models — you choose where the processing happens.

Prompt audit

History of AI interactions per user. Compliance with internal policies.

Honest considerations

AI quality depends on the chosen model. Smaller local LLMs have reduced capability vs. large corporate models.

The AI assists, doesn't decide. Suggestions are presented for the modeler to approve — never applied automatically without confirmation.

05 / AUTOMATIC DOCUMENTATION

The model becomes a document.

Technical documentation generated from the model itself — always up to date, in multiple formats.

Complete structure

Cover, table of contents, executive summary, data dictionary, diagrams, relationships, and business rules.

Bilingual EN/PT

Documents generated in English or Portuguese from the same model. Translation of descriptions supported.

Multiple formats

Export as PDF (distribution), HTML (web/internal portals), and PNG (presentations).

Customizable templates

Cover, header, footer, and visual identity configurable per organization.

Incremental generation

Update only modified sections. No need to regenerate everything on each change.

Version as artifact

Each generated document is associated with a model version. Full traceability.

06 / ON-PREMISE

Your data stays where it should stay.

Full deployment in the client's infrastructure, with real data sovereignty.

Deploy in your environment

Runs on your own servers or private cloud. No multi-tenant SaaS.

No external telemetry

No mandatory usage data sent outside the environment — not anonymized, not aggregated.

Flexible AI location

Use your corporate LLM, host local models, or disable AI — your choice by policy.

Containerized

Deploy in Docker / Kubernetes environments. Compatible with modern DevOps practices.

Backups and recovery

Models, metadata, and configurations in exportable format. Backup is the client's responsibility, using their own tools.

Controlled updates

Versions made available for deploy when the client decides. No forced updates.

MODULE 02

Data Governance

A practical application of DAMA-DMBOK® concepts: accountability guides metadata management, architecture, quality, and security. Explore how each discipline maps to 4dbAI features below.

About DAMA-DMBOK® at DAMA International
07 / CATALOG AND GLOSSARY DAMA-DMBOK® · Metadata management

Make organizational data easy to find and understand.

Discover assets across sources and give each term an agreed definition.

Connector-based discovery

Index metadata from relational databases, data lakes, and BI tools.

Searchable inventory

Find sources, tables, attributes, and descriptions in one catalog.

Semantic glossary

Corporate terms and approved definitions align business and technical vocabulary.

Business context

Link terms to data assets so meaning stays connected to usage.

08 / DATA LINEAGE DAMA-DMBOK® · Data architecture and integration

From source to consumption, every dependency in view.

Trace how data moves and assess the effect of a change before making it.

End-to-end traceability

Visual, automated mapping from origins to applications, dashboards, and reports.

Layered architecture

See data move through ingestion, transformation, and consumption.

Impact analysis

Identify tables, pipelines, and reports affected by a change at the source.

Change planning

Use identified dependencies to coordinate fixes with the teams involved.

09 / DATA QUALITY DAMA-DMBOK® · Data quality management

Clear rules, visible issues, accountable teams.

Monitor quality with measurable criteria and handle anomalies in context.

Completeness and uniqueness

Find missing values and duplicates that undermine processes and analyses.

Consistency and accuracy

Validate patterns, relationships, and compliance with organizational rules.

Anomaly dashboards

Track validation results and quality incidents in one place.

Targeted alerts

Notify technical and business owners to investigate and resolve issues.

10 / PRIVACY, LGPD AND ANPD DAMA-DMBOK® · Data governance and security

From source to evidence: data protection you can demonstrate.

Locate personal and sensitive data, follow its flows, and link every asset to owners, permissions, and records. Give controllers and data protection officers context to assess risk, prepare impact reports, and account for decisions when needed.

Regulatory context

Ongoing governance for a stronger ANPD.

Law No. 15,352/2026 established Brazil's National Data Protection Agency as a special autonomous agency with the independence to carry out its duties. The LGPD requires accountability; the ANPD may request information and impact reports under applicable law. 4dbAI helps keep the data, flows, and controls that support those responses organized.

  1. 01 · IdentifyClassify the data
  2. 02 · TraceUnderstand its flow
  3. 03 · ControlAssign access and ownership
  4. 04 · EvidenceSupport decisions and assessments

Inventory and classification

Identify and tag personal and sensitive data in the catalog. Use lineage to see where it flows and which assets depend on it.

Information for impact reports

Bring together data categories, flows, owners, and controls to help the controller assess risks and prepare a Data Protection Impact Assessment (RIPD).

Ownership and access

Assign Data Owners and Data Stewards to domains, link permissions to assets, and retain audit trails for relevant actions.

Accountability evidence

Review classification, lineage, and action history in context for internal assessments and to help respond to ANPD oversight requests.

Let's talk

Your models and data, with better context.

Live demo with your actual model. We respond within one business day.