100% Private — No Upload72,043 rows/sec14 Cleaning Operations

The Excel Cleaner That Lets You Verify It Never Uploaded Your File

14 cleaning operations. Fuzzy deduplication. Column profiling. Conditional rules. All browser-side — open DevTools and watch zero bytes leave your device.

72K+
Rows/sec
14
Operations
Zero
PII uploaded
None
Install needed

No account required · No upload · Any modern browser

You exported the file. It imported with errors. Finance says the vendor totals don't match. Now you're hunting invisible merged cells across 380,000 rows while the deadline ticks closer. You've been here before. You'll be here again — unless the cleaning happens before the import.

Why Excel Cleaning Breaks Pipelines

The file looks fine in Excel. Then it hits your import and everything breaks.

Merged cells silently corrupt imports
Only the top-left cell keeps its value. Everything else becomes blank. Your GROUP BY queries return wrong counts and nobody notices until the report ships.
Duplicate vendors with 3 different spellings
Excel's Remove Duplicates catches exact matches only. "Accenture" vs "Accenture Inc." vs "ACCENTURE INC" — all treated as separate entries, all skewing your totals.
Numbers stored as text break every formula
SUM() returns 0. AVERAGE() throws #VALUE!. It happens silently. You don't notice until the finance report has already gone out.
14 date formats in the same column
MM/DD/YYYY from US exports. DD/MM/YYYY from EU partners. YYYY-MM-DD from your database. ISO 8601 from your API. All mixed in one column.
Hidden rows from past imports nobody removed
Thousands of rows that are technically in the file but don't show in any filter. They break row counts and throw off totals invisibly.
Sensitive files cannot go to cloud cleaning tools
HIPAA, GDPR, or internal policy says no upload. But your workflow demands clean data. You're stuck between compliance and productivity — or you were.

Use Excel If... Use SplitForge If...

Stick with Excel if...
Your file is under 10K rows and you only need to remove a few blank rows
You need charts and pivot tables to stay dynamically linked after cleaning
Your team uses Excel macros for live formula-based transformations
You need tracked changes or version history built into the spreadsheet
Your workflow requires complex custom VBA automation
Use SplitForge when...
Your file has 100K+ rows and Excel is crashing or crawling
You need to catch near-duplicate vendors, contacts, or products — not just exact matches
Your data cannot be uploaded to any cloud service (HIPAA, GDPR, internal policy)
You need a reusable cleaning recipe applied consistently across monthly files
Your import is failing and you need a data health report before you fix it
You want to preview every change before committing — no irreversible steps

14 Operations. One Pass.

Everything a data analyst needs to make an Excel file safe to import — in a single browser session.

One-click deep clean

14 Cleaning Operations

Strip formatting, flatten formulas, remove merged cells, trim whitespace, normalize text case, remove empty rows/columns — all in one pass, no code required.

Catch near-duplicates

Fuzzy Deduplication

Levenshtein distance matching (0.50–0.99 threshold) plus Soundex phonetic matching. Catches near-duplicates that Excel's Remove Duplicates misses entirely.

Know your data first

Column Profiling & Health Report

Automatic data quality scoring per column — null counts, type distribution, outlier detection (3σ), top values, and numeric stats. Know what's broken before you clean.

Reshape without code

Split & Merge Columns

Split by delimiter, fixed position, or regex. Merge columns with custom separators. Keep or delete originals. No code required.

Smart data transforms

Conditional Rules Engine

Build IF/THEN rules with AND/OR logic — replace values, delete rows, or flag records. Equivalent to nested Excel formulas without the formula complexity.

CRM-ready in seconds

CRM & Contact Cleaning

One-click CRM preset: normalize phones, lowercase emails, proper-case names, fuzzy deduplicate contacts, expand abbreviations.

Designed for Sensitive Workflows

SplitForge processes files entirely in your browser using Web Workers and the File API. No bytes reach a server. This architecture is designed to support workflows where files cannot leave the device — including environments governed by HIPAA, GDPR, or internal data handling policies.

You don't have to take our word for it. Open Chrome DevTools, go to the Network tab, drop your file in, and you will see zero outbound data requests. The proof is auditable by anyone.

No server endpoint exists in this tool's architecture.

File API reads locally — no upload step exists in the architecture
Web Worker runs in a sandboxed thread with no network access
No analytics on file contents — only tool interaction events are logged
Verify yourself in DevTools: zero bytes outbound

How It Compares

FeatureSplitForgeExcel (manual)OpenRefineTableau Prep
Data upload required
Never
N/A (local app)
Local install
Yes (cloud)
Install required1
No — browser
Yes
Yes (Java)
Yes
Max rows before slowdown3
Tested reliably to 1M+ rows
~100K
~500K
Server-limited
Fuzzy deduplication1
Yes (Levenshtein)
Exact match only
Yes
Limited
Column profiling / health1
Yes (auto)
Manual (formulas)
Basic
Yes (paid tier)
Conditional rules engine
Yes
VBA / manual
GREL scripting
Yes (complex)
CRM preset (1-click)
Yes
No
No
No
Merged cell handling
Auto-fill + remove
Manual
N/A
Limited
Date normalization
Auto (14 patterns)
Manual or formula
Via GREL
Yes
Privacy by architecture2
Yes
Yes (local)
Yes (local)
Contractual only
Workflow export (reusable)
Yes (JSON export)
Macro (.xlsm)
Yes (.json)
Yes (.tfl)
Before/after preview
Yes
No
Yes
Yes
No technical knowledge needed1
Yes
Partially
No (GREL)
Partially

1 Install requirements and feature availability sourced from G2 reviews and vendor public documentation, Feb 2026. 2 Tableau Prep cloud upload applies to Tableau Cloud (formerly Tableau Online); Tableau Desktop processes locally. 3 OpenRefine row limit estimates based on community benchmarks and project documentation. SplitForge data from internal testing — see the full performance page for methodology.

Ready to run it on your file?

Drop it in. Get a health report. Clean in one pass. No account needed.

Time Savings Estimate

Enter your current manual cleaning workload to estimate the time difference.

4
2
$55
8.0h
Manual hrs/month
8.0h
Hours saved/month
$438
Saved/month
$5,251
Saved/year

Assumes ~40 seconds per file (100K-row standard clean). Actual time varies by file size, complexity, and operations selected.

What This Looks Like in Practice

Representative workflow scenarios — not customer case studies. Times reflect mixed operation sets including fuzzy matching; analysis-only speeds are higher.

Finance · Representative scenario

Quarterly Vendor Reconciliation

Before

380K-row vendor ledger from SAP — merged cells in every group header, multiple date formats across fiscal year columns, vendor name variations (e.g., "Accenture" vs "Accentur" vs "ACCENTURE INC.").

Manual cleanup: 6–8 hours in Excel
After

Merged cells removed, dates normalized to ISO 8601, fuzzy deduplication (0.85 threshold) collapses near-duplicate vendor entries, business suffixes standardized.

~47 seconds in SplitForge · 380K rows
Healthcare · Representative scenario

Patient Registry Normalization

Before

520K-row registry with phone numbers in 8+ formats, addresses as single strings, mixed data types across 34 columns, hidden rows from previous imports.

Manual: 2–3 days across multiple analysts
After

Phones normalized, addresses parsed into structured fields, numbers-as-text converted, hidden rows surfaced. Processed without upload — required for PHI workflows.

~72 seconds in SplitForge · 520K rows
E-commerce · Representative scenario

Multi-Supplier Product Catalog Merge

Before

750K-row catalog from 4 supplier feeds — inconsistent column headers, prices with embedded currency symbols, exact and near-duplicate product entries.

Manual: 4–5 hours with VLOOKUP + review
After

Headers normalized, currency symbols stripped, exact and fuzzy duplicates removed, conditional rules flag items with $0 price for pre-import review.

~38 seconds in SplitForge · 750K rows

Technical Deep Dives

How the hard problems are actually solved.

When SplitForge Is Not the Right Tool

Browser memory cap: ~1GB for Excel files (varies by OS and available RAM). Files over 500MB: close other tabs first. Files over 1GB: use Excel Splitter to process in chunks.
Fuzzy deduplication slows significantly above 100K rows — O(n²) algorithm; expect 2–15 minutes for 500K rows. Run exact dedup first to reduce dataset size before applying fuzzy matching.
No API or CLI — cannot run in automated pipelines, cron jobs, or CI/CD workflows. For batch automation, use Python + openpyxl/pandas.
Password-protected workbooks not supported.
Formulas involving external workbook references cannot be flattened.
Charts and embedded images are preserved but not modified or cleaned.
ODS, Numbers (.numbers), and Google Sheets formats not supported. Convert to XLSX first using Format Converter.

For automated pipelines, use Python + openpyxl/pandas. For large-scale fuzzy clustering, use OpenRefine. For enterprise ETL orchestration, use Tableau Prep or AWS Glue.

File too large to process in one pass? Excel Splitter breaks it into chunks first. Need to mask sensitive columns before cleaning? Data Masking uses the same browser-only architecture. Working with CSV instead of Excel? CSV Data Cleaner covers the same 14 operations for flat files.

Performance Overview

100K500K1M0s4s8s12s16s

Filled bar (1M rows): Verified — Chrome stable, Windows 11, Intel i7-12700K, 32GB RAM, February 2026. 10 runs, highest/lowest discarded, remaining 8 averaged. Analysis only (no cleaning operations applied).
Semi-transparent bars: Calculated from 72,043 rows/sec baseline. Results vary by hardware, browser, and file complexity (±15–25%).

Frequently Asked Questions

Stop Cleaning Manually. Start in 10 Seconds.

Drop your file. Get a health report. Clean in one click. Download a file you can actually import.

14 operations — merged cells, fuzzy dedup, column profiling, conditional rules
100% private — file contents never leave your browser
Before/after preview — see every change before committing
Workflow export — save and reuse your cleaning recipe

Also try: Excel Splitter · Data Masking · Remove Duplicates · Data Validator