These are illustrative scenarios showing how the products are used, not customer case studies.
The same sequence, twice
Today
- Customer sends a file
- Importer grows a special case
- Load fails after submit
- Rows repaired by hand
With SynaptaGrid
- Send them the sample file
- Import CSV or JSON
- Model rules reject bad rows
- Duplicate rules flag repeats
The situation
The columns are named differently, the dates come from three locales, and some of the rows already exist under a slightly different spelling. The importer grows a special case per customer, failures are found after the load rather than before it, and someone ends up repairing the result by hand.
How it is approached
Every published version can be imported and exported as CSV or JSON, with a sample file showing the shape it expects. The model’s own rules — required, unique, typed, validated — apply on the way in rather than being restated in an import script. What counts as a duplicate is defined on the schema, with a similarity threshold for the near-misses, and those rules are versioned alongside it. The loaded records land in the generated screens, where the person who owns the data can filter and fix them.
What changes
Loading a customer’s data stops being a script per customer. The rules that reject a bad row are the same rules the API enforces afterwards, so an import cannot introduce something the product would refuse.