erwin Data Modeler Generative AI: Data Sharing, Privacy and Protection Methods - FAQ

Category

What erwin DM Delivers

AI Modeling Capabilities

AI leverages large language models (LLMs) to transform plain-language requirements into structured data models. It can automatically generate entities, attributes, relationships, metadata, business terms, and DDL scripts.

Integration & Modernization

Supports SQL, NoSQL, Snowflake, BigQuery, Databricks, MongoDB. Recommendations for migrations and erwin ER360 integration are planned.

Authentication & Access Control

Access via SSO, LDAP, or Entra ID. AI usage is managed per license and can be disabled or blocked at the organization level.

Encryption & Data Security

Data is encrypted in transit (TLS 1.2+) and at rest (AES-256, FIPS 140-2). No data is stored by erwin DM.

Prompt Privacy & Data Handling

Prompts and responses are not stored externally. Feedback submissions are optional. Microsoft’s terms protect against training use.

Infrastructure & Hosting

AI models are hosted in the U.S. East region. Data residency control is not available. The Service level agreement (SLA) uptime is 99.90%.

Governance & Compliance

Compliant with GDPR and CCPA. Prompts are never used for training. AI does not auto-detect sensitive data.

Explainability & Trust

AI does not currently explain its suggestions. Human validation and transparency controls are planned.

Future Roadmap

Upcoming features include conversational modeling, natural language queries, and Agentic AI for enterprise-scale governance.

Frequently Asked Questions (FAQ)

erwin DM uses the GPT-4 base model, deployed within Microsoft Azure.

No, prompt data is not stored or reused by OpenAI.

Yes, only the user’s email address and license number are stored.

All data processing takes place in U.S. regions within Quest’s Azure environments.

Yes, it can be disabled.

It is powered by GPT-4.

Because OpenAI does not use this data to train its models.

You can download the PDF version of this guide here.