Customer guide

Getting started

Upload a file, inspect preflight, confirm mappings and settings, process it, then review results before downloading.

Supported formats

CSV and XLSX work best, including multi-sheet XLSX. DOCX needs structured tables; PDF needs machine-readable tables. Scanned PDFs are unsupported.

File preparation

Use one header row and distinct column names. Store identifiers as text when leading zeroes matter. Remove presentation-only rows when practical.

Sales data

Common fields are sales record ID, amount, sale date, status, customer, and notes. Preserve business-specific text when no mapping applies.

Leads / CRM data

Common fields are lead ID, name, company, email, phone, created date, and lead status. Email and phone are optional unless you require them.

HR data

Common fields are employee ID, name, department, designation, joining date, and employment status. Review unknown statuses. Never use output for automatic employment decisions.

Column mapping

Review the suggestion, confidence, and samples. Confirm uncertain high-risk mappings, choose another supported type, or Preserve as-is.

Required fields

Choose only fields essential to your use case. Missing or invalid required values can block Ready to Use; optional blanks do not automatically block.

Cleaning settings and rules

Toggle only supported safe normalizations. Recognized names, emails, phones, dates, and statuses can be normalized. Ambiguous dates and unknown statuses are preserved rather than guessed.

Validation findings

Findings identify the row and column, original and cleaned values when available, severity, blocking state, and a review action.

Quality tiers

Ready to Use has sufficient quality and no blocking finding. Needs Review covers uncertain or correctable findings. Poor Quality indicates significant missing or invalid information.

Issue rows

Issue Rows counts unique records affected by validation or anomaly findings. Multiple findings on one record still count once.

Duplicate candidates

Candidates may match exactly or fuzzily. Review matching and conflicting fields; records are never automatically deleted or merged.

Column analysis

Mapping confidence, missingness, invalid formats, changes, and required status explain a column. One uncertain column does not make the whole dataset bad.

Downloads

Browser tables are previews or paginated collections; generated dataset downloads contain the complete processed output. Start with Cleaned CSV, Datario Excel Report, and Findings Summary.

DOCX limitations

Only structured Word tables are supported. Review extracted results before operational use.

PDF limitations

Only machine-readable tables are supported. Extraction may alter formatting, so reconcile results with the source.

Privacy and retention

Temporary uploads are removed after each request. Run artifacts become eligible for periodic cleanup after approximately 24 hours; exact-time or secure deletion is not promised.

Troubleshooting

If inspection fails, check that the file is not empty, damaged, password-protected, or missing a supported table. Retry once, or reset and upload CSV or XLSX.

Frequently asked questions

Which formats work best?

CSV and XLSX.

Does Datario delete duplicates?

No. It reports candidates for review.

Will my original file be changed?

No. Processing creates new artifacts.

Can Datario process scanned PDFs?

No; OCR and image-only PDFs are unsupported.

Why must PDF results be reviewed?

Extraction can alter layout or formatting.

What does Ready to Use mean?

The record meets the quality threshold and has no blocking finding.

What does Needs Review mean?

The record has an uncertain, correctable, anomaly, or duplicate-related finding.

What counts as an Issue Row?

One unique record affected by validation or anomaly findings.

Can I change detected mappings?

Yes, using a supported type or Preserve as-is.

Can I choose required fields?

Yes, per dataset.

What happens to unknown statuses?

They are preserved and flagged.

Are all rows shown in the browser?

No. Browser views are limited or paginated.

Are complete rows included in downloads?

Yes, complete generated datasets are downloaded.

How long are files retained?

Artifacts become eligible for periodic cleanup after approximately 24 hours; timing is not exact.

Can Datario make employment decisions?

No. HR output requires human review.

What should I do if a file fails?

Check its structure, retry, or upload a CSV or XLSX copy.