Validate a CSV before import
Check required fields, formats, codes and duplicates against agreed rules. Separate rows that can proceed from rows needing review.
For SMEs, agencies and administrative teams
If you download a file, fix the same errors and rebuild the same document every week, that step can be specified and automated. Ospivia builds dedicated data, reporting and document workflows around an agreed input format and output.
An export with ten columns is not yet a specification. A date can mean day/month or month/day; a code with leading zeros is not a number; a repeated row can be an error or a separate transaction. Automation should expose these rules and stop ambiguous cases rather than produce an apparently clean file.
Check required fields, formats, codes and duplicates against agreed rules. Separate rows that can proceed from rows needing review.
Group compatible exports, compare equivalent periods and show excluded anomalies. First define returns, cancellations and the meaning of each total.
Group by order, validate quantities and amounts and generate consistent PDFs. An invalid row can block the whole order according to an agreed rule.
First Result · EUR 790
The EUR 790 First Result offer can cover one input schema, an agreed set of rules and a defined output. Starting with CSV limits dependencies. Excel workbooks, macros, formulas or multiple sheets need a specific assessment: the linked demonstrations process CSV files.
EUR 200 at the agreed start and EUR 590 after testing and written acceptance. First delivery within 10 business days of the agreed start, with materials and access ready. Eligibility, date, total including applicable taxes and external costs are confirmed before the order.
Business-system access, unlimited archive cleaning, two-way synchronisation, email delivery, server scheduling and tax documents are not automatically included. Frequency, volume, execution and error handling belong in the initial assessment.
Read the complete scope and guaranteeThese are synthetic demonstrations, not customer results. Each page explains what ran, what you can download and what the example does not prove.
14 synthetic rows: 6 valid and 8 needing review. Compare the files and the check log.
22 synthetic rows: 16 accepted, 1 duplicate and 5 quarantined. The report also exposes what is excluded from totals.
10 rows, 3 orders: 2 PDFs generated and 1 entire order blocked. The actual outputs are available to download.
Every input row should be accounted for: accepted, excluded or needing review. Counts and, where applicable, amounts must reconcile.
The same inputs and configuration should produce consistent content. Document any variable fields, such as the generation timestamp.
An unexpected schema or ambiguous value should produce a useful message instead of silently replacing a valid output.
Describe the current steps, tools, input, expected output and acceptance test. Our plain-text template includes a fictional example and can be edited in any text editor. No registration is required.
Use fictional examples. Do not include passwords, personal data or confidential documents in the initial enquiry. Nothing is submitted by downloading the template.
No. Column names and a few invented rows are enough for an initial assessment. If real data are needed later, we agree the contents, access and sharing method.
Only if scheduling, an execution environment and failure handling are included in the scope. The public examples process files; they are not already connected to your accounts.
An agreed export and import can often work alongside your current tools. The actual supported schema must be checked; we do not promise an integration before verifying it.