June 20, 2023

Conceptual Model Conversion - Ellie Version 4.7 Release

Data Modeling
Product Design
Juha Korpela
Chief Product Officer, Ellie.ai

So you’ve created a nice conceptual model to describe your business data needs, and you want to take this a step further towards implementation design by creating a logical model? That just became a lot easier with Ellie v4.7!

Our new feature, Conceptual-Logical model conversion, allows you to create new logical models directly from existing conceptual models. This is done with a simple “Convert” button on the conceptual modeling canvas.

When the “Convert” button is clicked, Ellie automatically creates a new logical model with matching entities, relationships, and attributes:

Each logical entity gets its name & attributes from the source conceptual entity, and relationships are created with the exact same cardinalities. Naturally, all of these can be then modified in the new logical model as you see fit! Because it’s important to always know what your logical entities are, we also create links between the new logical entities and their conceptual sources.

Using model conversion to speed up development

When you want to ensure that your development efforts are going in the right direction, a good data model is essential. Logical models are good design documentation for your data products, but it’s important that they are based on real-life business. The best practice here is to document your business requirements with a conceptual model, and from that derive logical models for different use cases.

With Ellie’s new model conversion feature, you get a first-draft version of the logical model immediately, based on your conceptual model. You can then modify this logical model to fit your needs: rename entities into facts & dimensions if you’re building a star schema, design hubs & links if you’re building a Data Vault, et cetera.

Remember that you can always create another logical model from the same conceptual model - as many as you like! This allows you to use the conceptual model as an overall guide of “what should be”, and utilize different logical models for different kinds of technical needs. All the models can be exported from Ellie and linked with various data engineering tools and data catalogs to drive your implementation.

Other changes in v4.7

The latest Ellie version also brings some minor changes elsewhere. For those of you who are integrating Ellie’s Glossary with your data catalog solution, we have added an “updated at” field into the entity metadata available via our Glossary API endpoints. This helps you find the latest changes when e.g. synchronizing content between Ellie and a data catalog.

Regarding integrations, we are also releasing another example integration solution in our open-source GitHub repository. The latest addition there is a JavaScript solution for importing logical models from PostgreSQL and Snowflake; a Python solution for the same was already added earlier. You’re free to use these as the basis of your own integration solutions, or as-is!

Now go check out the new version!