Generating a Model Using AI
Go Up to Creating and Working With Data Models
Note: This feature is currently Public Preview and is planned for a formal release on October 31, 2025. As a Public Preview, it is fully tested and free of charge. We welcome your feedback. The feature is OFF by default and requires you to accept the terms and conditions before use.
WARNING: The accuracy of data models generated using AI is not guaranteed. Any content generated using this tool must be separately reviewed and verified by a human.
Note: At the time of initial release, ER/Studio Data Architect used OpenAI's GPT-4.1 to generate the model data for this feature. This may change at any time.
In addition, ER/Studio Data Architect data model generation is powered by GPT models that enforce a maximum token output setting. This feature is not intended to build large, complete logical models. User requests that result in model data that exceeds the maximum number of tokens may result in incomplete or failed responses.
To use this feature, make sure that the host aida-api.idera.com is whitelisted in your network firewall settings. To prevent users from accessing this feature, block aida-api.idera.com in your network firewall settings.
Contents
Getting Started
ER/Studio Data Architect 20.8 and later allow you to create a logical data model using a block of text. For general questions about this feature, see AI Model Builder FAQs.
About AI Model Generation
ER/Studio Data Architect 20.8 and later allow you to create a logical data model using a block of text. This feature is available by clicking Generate Model using AI when creating a new file or by clicking the Generate Using AI button in the Tools ribbon bar, as shown in the following image. Note that the Welcome Page also offers a link to generate a model using AI.
The AI Model Builder dialog allows you to type prompts for the tool to create a new model. Click Surprise Me to include an example prompt so that you can get an idea of what information and format helps the AI build a good model. The following example was generated using the Surprise Me button and explains how to build a database to manage a large-scale event planning and ticketing platform.
Using AI to Generate a Model
To generate a data model using AI
- Use one of the following methods to access the AI Model Builder:
- Click File > New and then select Generate Model using AI before clicking OK.
- Click AI Model Builder in the Tools ribbon bar.
- On the Welcome Page, click Generate Using AI.
- If this is your first time using this feature, review the displayed terms and conditions for using the Data Architect AI model builder.
- If you acknowledge and agree to the terms, click Accept. Note that you must agree to use this feature. You can access the terms and conditions any time after accepting by clicking the Terms button in the AI Model Builder window, which appears after you click Accept.
- In the Model Prompt field of the AI Model Builder, input a description of the model that you want to build. Note that the Surprise Me button provides an example prompt so that you can get an idea of what information and format helps the AI build a good model. You can always change or delete text typed in this field.
- In the Options area, select the appropriate initial layout. Hierarchical is selected by default.
- Optional. Click Remember this text to save the entered prompt text to appear the next time you open this dialog.
- Click Generate. The AI Model Builder displays a progress dialog while building the model based on your input.
- When finished, the AI Model Builder displays your new model in the background while also displaying a robust description of the model in the AI Model Builder Progress dialog.
- Click Save to File or Print to retain a copy of the log generated during model creation.
- Click Close to close the AI Model Builder Progress dialog.
Supported Objects
The following objects and properties are potentially available in this feature:
I. Entities
A. Name B. Description C. Attributes 1. Name 2. Data Type i. INTEGER ii. DATE iii. DATETIME iv. VARCHAR v. CHAR vi. DECIMAL vii. BIT 3. Primary Key Flag 4. Foreign Key Flag 5. Alternate Flags 6. Null/Not Null property 7. Width and Scale 8. Description
II. Relationships
A. Normal 1. Non-Identifying/Identifying Flag 2. Forward verb phrase 3. Reverse verb flag 4. Cardinality 5. Role Names B. Super/Sub 1. Discriminator property 2. Type i. Complete ii. Incomplete 3. Membership Type i. Exclusive ii. Inclusive 4. SubTypes
III. Title Block
A. Contains the summary of the results