An intention. A journey. A tool.
V / 08 — Data & connections
Check data before bringing it in.
START WITH THE SITUATION
A POSSIBLE USEExample: bring a list of ordinary references and categories into an existing tool, flagging empty references and duplicates.
A missing reference, malformed date or duplicate row can disrupt a tool. We build an input check for a specific format: read the file, verify columns, identify rows to correct and show a preview before import.
Permitted transformations are documented. The journey shows what is accepted, what needs review and what has actually been transferred to the agreed destination. The starting price covers one small, stable CSV format rather than a universal data-processing platform.
This simulation checks references and duplicates across three simple columns, without quotes or line breaks inside a cell.
Local demonstration · fictional data · no real operation.
THE STARTING PACKAGE
* Tax-inclusive starting price and indicative duration for this scope. A written proposal follows assessment, with total, dependencies and third-party costs before commitment.
1 stable CSV format, up to 8 columns and 500 rows.
Up to 6 simple validation rules.
1 preview screen and 1 agreed output.
1 language, without sensitive-data storage.
THE AGREED RESULT
A format check and row preview.
Error messages linked to the relevant fields.
A specification of the accepted file and intended handling.
THE DELIVERY PATH
Set the format and acceptance rules.
Create reading, validation and preview.
Test invalid rows and the agreed output.
BEFORE YOU CHOOSE
Only for transformations defined in advance. Other errors are flagged for a decision; arbitrary corrections should not be applied.
Behaviour depends on the destination. A preview is not a backup; a rollback method must be arranged before any actual write.
THE RIGHT NEXT STEP
A context, a constraint, an idea. Let’s use that to choose a realistic first release.
Prepare my project