Free · Mauritius e-invoicing · Data quality
Is your invoice data ready for e-invoicing?
Upload invoice exports and check each field against the published MRA e-invoice field rules: tax codes, buyer types, VAT numbers, BRNs, date formats and totals. Every finding says what's wrong, where it is, and how to fix it.
- 33 checks, each with a rule id and its official source
- Exact cent-level arithmetic on VAT and totals
- Excel and PDF report you can hand to your accountant
Private by design. Files are processed in memory and discarded as soon as your results are returned. Nothing is stored, no account is needed, and invoice contents are never logged or sent to any AI service. We only count anonymous totals (checks run, rows checked, which rules fail most).
This is a data-quality check, not tax advice.
Demo workspace
Want the full demo?
Ask for your own demo login. You get a workspace with a full batch of made-up invoices, every finding, and the Excel and PDF reports. The free check above stays open to everyone.
Already have a login? Sign in to the demo.
Results
Your invoice health check
Most common issues
Click a rule to filter the invoices.
Email me the full report Coming soon
Not switched on in this preview. Nothing you type here is sent or saved. Use the download buttons above.
No invoices match these filters.
Fictional sample invoices
Nine made-up invoices. Each one hides a known mistake, and one is clean. Download them, check them here, then compare with the answer key.
| Sample | Seeded issue | Files | Check |
|---|
How it works
Export
Export invoices and invoice lines from your billing or accounting system as CSV or Excel. Zip a batch if you have several files.
Check
Each invoice and line runs through 33 field checks taken from MRA's published e-invoice data structure and templates.
Fix
Drill into every finding down to the file, row and field, with plain-English fix guidance. Download the report for your team or auditor.
Why not just paste invoices into ChatGPT?
General-purpose AI chat is great for many things. A compliance data check needs a few properties it isn't designed to guarantee.
Rules from the published spec
Every check comes from MRA's published e-invoicing documents. Each finding carries a rule id, its source document, and whether the rule is confirmed or the templates disagree.
The same answer every time
The checks are fixed code, not a generated reply. Run the same file twice and you get the same findings, which is what you want before you change a billing system.
Exact cent-level arithmetic
VAT at 15%, line totals and header totals are recalculated with decimal arithmetic and rounded to the cent, not estimated.
Client data stays out of AI tools
Files are processed in memory and discarded. No invoice data is sent to a third-party AI service, so you're not sharing client names, VAT numbers or BRNs with anyone.
Whole batches, structured output
Upload full Excel exports or a zip of files. Results come back per invoice, per line and per field, with file and row numbers, not as a paragraph of text.
A report you can show an auditor
Download an Excel or PDF report listing each check, its result, its source and how to fix it: a clear trail of what was checked and when.
Limits and what this is not
- Up to 4.4 MB per check, 10 files (or a zip of up to 20 CSV/XLSX files), 5,000 invoices and 50,000 lines.
- It checks field formats, allowed codes and arithmetic. It does not fiscalise invoices, connect to MRA, obtain an IRN or QR code, or decide how a supply should be treated for VAT.
- Where the two official MRA templates disagree, the finding is marked conflict; hints from other public documents are marked unverified and never fail an invoice on their own.