Sift

Get a messy CSV import-ready for your CRM, without it ever leaving your browser.

Sift is a free, open-source data cleaner for CRM imports and migrations. Drop a CSV or Excel file and it dedupes, normalizes, and maps your data to the schema you're importing into, then checks it will actually import. Deterministic and audit-first: you approve every change, and nothing is uploaded.

Loading the tool. It runs entirely in your browser; your data never leaves your device.

What Sift does

  • Clean the usual mess: stray spaces, inconsistent casing, placeholder blanks, broken emails, mixed dates, currency text.
  • Normalize phone numbers (E.164-style), countries, and UK/US/Canadian postcodes.
  • Dedupe exact and fuzzy duplicates with survivorship rules to build one golden record per person.
  • Reshape: combine or split columns, look up and join from a second file, filter rows, tag a whole list.
  • Score and qualify leads by persona, then extract the segment you want.
  • Map columns to your CRM's import template and pre-flight whether the file will actually import.
  • Export a clean file, a deduped accounts/companies file, or a hygiene report.

Who it's for

  • RevOps and CRM admins running a migration or import (Salesforce, HubSpot, Dynamics, Zoho, Pipedrive).
  • Field and event marketers prepping a lead or attendee list after a conference.
  • Sales and data ops cleaning and deduping contact and account lists.
  • Privacy- or compliance-constrained teams (legal, healthcare, finance, EU/GDPR) who can't paste customer data into an online tool or an LLM.

Why Sift

  • Private by architecture. Runs entirely in your browser, nothing uploaded. The opposite of pasting your list into ChatGPT or a web cleaner.
  • Import-readiness, not just cleaning. Maps to your target's schema and checks it will actually import, with one-click fixes.
  • Deterministic and auditable. Every change is a rule you approve with a before-and-after diff and an audit trail. No AI black box, no account.

Frequently asked questions

What is Sift?

Sift is a free, open-source, browser-based tool that cleans a messy CSV or Excel file and gets it import-ready for a CRM, entirely on your device.

Is my data uploaded anywhere?

No. Sift runs 100% in your browser; your rows never leave your device and there is no server to send them to.

How is it different from cleaning data in ChatGPT?

Sift is deterministic and private: every change is a rule you approve with a visible before-and-after diff, nothing is sent to an AI model, and your data stays on your machine.

Does it work with HubSpot, Salesforce, and Dynamics?

Yes. Map your columns to the destination's import template and Sift checks required fields, types, and allowed values before you import. It also generates an accounts file for CRMs that need accounts created first.

Can it dedupe contacts?

Yes, exact and fuzzy dedupe matching on email, phone, or a fingerprint, with survivorship rules for one golden record per person.

Is it free and open-source?

Yes to both. No account, no paywall, no collection of your file's data.

How big a file can it handle?

Up to about 150MB in standard mode, and a large-file mode powered by in-browser DuckDB for hundreds of thousands of rows.