The iSynth approach:
treat data as code
Define data model
In iSynth, data is described in code. This starts with the data model, with an object-oriented approach. The key here is to deliberately keep the data model application-neutral and simple, and to focus only on business-relevant aspects. This enables good maintainability over time.
Define constellations
and variance
Data constellations are built on top of the data model by combining object types like Lego® bricks into an object graph. In doing so, we draw inspiration from achetypical data cases from the real world.
And since we also describe constellations as code, arbitrarily complex entities can be built algorythmically.
Variance can be generated via randomness or injected specifically from external sources.
Synthesize data
Synthesizing data feels like shopping: just select from available data constellations, define the quantity you need, and run the data order.
Now, all the magic happens in the background, and synthetic data is generated in the quantity and variance you defined.
Enrich and provision data
At this point, we have synthetic data, but in an application-neutral form. Now the next step is to get the individual data objects into a technical format that the target applications can load, using any appropriate application interfaces at hand.
iSynth comes with its own workflow engine to manage such data provisionings and offers a powerful API to automate the complete process, for example inside a CI/CD pipeline.
Learn more about iSynth
Do you still have questions? We love to tell you more about the iSynth approach and how you can improve your process with synthetic data.