What should a CSV import tool for SaaS include?

An embeddable widget your users recognize, column mapping that handles renamed and multilingual headers, validation with a review step, and reliable delivery into your backend — plus an audit trail for support. Formats beyond CSV (Excel, JSON, XML, PDF) remove the 'please re-export as CSV' support ticket.

Build or buy a CSV importer?

A file input and a parser take a week; the long tail — encodings, delimiter dialects, duplicate headers, header detection, mapping UX, validation, undo, audit — takes quarters. Buying makes sense when the tool embeds cleanly and doesn't force customer data through a third-party cloud. Mildport is built for exactly that trade: embeddable widget, self-hosted engine.

Can it run without sending data to a vendor?

Yes. Mildport self-hosts with Docker Compose or Helm — files stay in your VPC — and its licenses verify offline. There is even a zero-backend widget mode that parses, maps and applies entirely in the browser.

CSV import tool for SaaS — embed once, accept anything

Esta página aún no está disponible en tu idioma.

A CSV import tool for SaaS is the difference between “please format your spreadsheet like our template” and customers just… importing. Mildport embeds as one custom element, accepts what customers actually send, and — unlike the cloud importers — runs in your infrastructure.

The integration, honestly

npm install @capitality-io/mildport-widget

<mildport-import
  api-base-url="https://imports.your-infra.example"
  license-key="SIGNED_TENANT_KEY"
></mildport-import>

Attributes in, DOM events out — no iframe, no SDK runtime wrapping your app. Define target fields in code, in the no-code catalog editor, or draft them from your API spec. Listen for import-applied. The Quickstart is ~10 minutes.

What your users get

One wizard for every file — CSV/TSV in any dialect, XLSX/XLS/ODS, JSON, XML, Markdown and HTML tables, PDF tables, OCR’d scans and photos.
Mapping that survives reality — renamed headers, six languages, duplicate columns, header rows that aren’t row one. Deterministic scores, visible reasons.
Cleanup without a spreadsheet round-trip — plain-language fixes, previewed as before/after diffs; accepted fixes replay free on the next similar file.
A review step that catches problems before they land — typed validation, required fields, join across multiple files.

What your team gets

Per-tenant learned mappings that make the second import automatic; an audit surface for everything the engine learned; delivery via signed webhooks with retries; a decision log per import — and published, transparent pricing, including flat-annual self-hosting.

Next: try the playground with a messy sample file, read what is Mildport?, or embed via the CSV import SDK.