Pivot & Unpivot Massive CSV Files.
No Python. No Power Query. No Uploads.
Transform millions of rows into aggregated summaries, or reshape wide spreadsheets into normalized long-format tables — with 12 aggregation functions, wildcard/regex column selection, and zero file uploads.
Stop rebuilding the same pivot table every Monday.
You've been here before
Excel Pivot Tables Break at Scale
Hard row cap: 1,048,576. Performance typically degrades well before that. No true MEDIAN, no Count Distinct in standard Pivot Table UI (requires Data Model / Power Pivot), no StdDev natively. Unpivot requires Power Query — a separate wizard with manual column-by-column selection.
Python Requires Setup and Code
pandas.pivot_table() and melt() are powerful — but they need a Python environment, package installs, and someone who can write the code. RAM-bound: MemoryError onset varies by machine and dataset. Not accessible to the analyst who just needs the report done.
Cloud Tools Want Your Data
Power BI requires cloud upload to share results. Most browser CSV tools upload files to their servers to process them. Patient records, financial data, employee lists — those shouldn't go to a third-party server to get a pivot table.
Two Operations. One Tool.
Most tools do one or the other. SplitForge does both — in the same interface, with the same privacy guarantees.
How to pivot a CSV file without Excel or Python
Browser-based pivoting means the file stays on your machine — no upload step, no cloud processing. Upload your CSV or XLSX, select which columns to group by (e.g., Region, Product), pick your value columns (e.g., Revenue, Qty), choose aggregation functions (Sum, Count, Average, StdDev, and more), and click Process. For unpivot, select your identifier columns, then use a pattern like Q* to select all quarterly columns at once. Results download as CSV, JSON, TSV, or JSONL. For related operations, see SQL-style GROUP BY on CSV files and why Excel and Pandas crash on large pivots.
Everything Excel Pivot Tables Can't Do
Six capabilities that separate this from Excel add-ins and generic browser tools.
12 Aggregation Functions
10 memory-safe: Sum, Count, Average, Min, Max, Count Distinct (unique values per group, capped 100K), Weighted Average, % of Total, StdDev (Welford's O(1)), Variance. Plus 2 advanced: Median and Mode (behind explicit memory-warning toggle).
Wildcard + Regex Column Selection
Simple mode: Q* matches Q1, Q2, Q3, Q4. 2024_* matches all 2024 columns. Regex mode: ^Q[0-9]+ for stricter patterns. Patterns add to existing selection — mix patterns and manual clicks. Case-insensitive. Critical for 50–120+ column wide files.
3-Level Hierarchical Pivot
Create nested groupings: Year → Quarter → Month, or Country → State → City. Each level gets its own subtotal row. Levels are drag-and-drop reorderable. Available columns auto-filtered per level to prevent duplicates.
Config Save / Load / Share
Save up to 20 configurations with custom names (stored in localStorage — never leaves your device). Load any saved config for recurring reports. Export config as JSON to share. Import JSON configs from teammates. Applies to both pivot and unpivot modes independently.
Native XLSX with Sheet Selector
Upload Excel files directly. Multi-sheet workbooks show a sheet selector with row count previews. Single-sheet workbooks auto-select. Output to CSV, TSV, JSON, or JSONL — no forced XLSX output lock-in. XLS format also supported.
Streaming Architecture (10M+ Rows)
Web Worker processes data off the main thread — UI stays responsive throughout. 50,000-row chunks. Pivot: memory grows with group cardinality. Unpivot: memory stays flat regardless of file size. Tested at 10M rows (~18 sec, Feb 2026). Results vary.
How SplitForge Compares
Side-by-side: row limits, aggregations, unpivot method, pattern matching, privacy, and cost.
| Tool | Row Limit | Aggregations | Unpivot | Pattern Select | Privacy | Cost |
|---|---|---|---|---|---|---|
| Excel Pivot Tables | Cap: 1,048,576 | 11 (no MEDIAN) | Power Query (manual) | Local (desktop) | Microsoft 365 | |
| Power BI | Varies by tier | DAX (complex) | Power Query | Cloud upload | $14/user/mo | |
| Python pandas | RAM-bound | All (code required) | melt() (code) | Code only | Local | No install required |
| SplitForge← You are here | 10M+ (browser mem) | 12 incl. StdDev, CD | GUI + wildcard/regex | 100% client-side | See splitforge.app |
Real-World Use Cases
Six scenarios where Excel or Python fails — and how SplitForge handles them.
Sales Analytics Dashboard
800K sales transactions. Need monthly revenue by Region and Product with SUM, COUNT, AVG, and StdDev. Excel Pivot Tables degrade near this size and lack StdDev in Pivot Table UI.
Pivot by Region, Product, Month. SUM(Revenue), COUNT(Orders), AVG(Revenue), StdDev(Revenue). All memory-safe aggregations — no buffering. ~8 seconds total.
Financial Reporting Normalization
Budget spreadsheet with 36 month columns (Jan_2024 through Dec_2026). Need Date and Amount columns for Power BI. Excel Power Query requires 8+ manual steps with no pattern matching.
Wildcard *_202* selects all 36 month columns instantly. Output: Department, Category, Month, Budget. 12 seconds vs. 20+ minutes manual Power Query.
Healthcare Patient Metrics
1.2M patient visit records. Analyze outcomes by Hospital → Department → Procedure with average stay, Count Distinct patients, and cost variance. Compliance-constrained — no cloud uploads.
Hierarchical pivot: Hospital/Department/Procedure. AVG(Stay Days), Count Distinct(PatientID), Variance(Cost). 100% client-side. ~42 seconds.
E-commerce Catalog Restructure
15K products with separate columns per marketplace: Amazon_Price, Amazon_Stock, eBay_Price, eBay_Stock, Walmart_Price, Walmart_Stock. Need normalized rows for database import.
Regex patterns Amazon_*, eBay_*, Walmart_* via three pattern-match runs. Result: 90K rows (15K × 3 × 2) in normalized format. 8 seconds.
Financial Portfolio Analysis
2M investment transactions across 5K accounts. Portfolio performance by Account Type → Region → Asset Class. Need Variance and StdDev. Excel lacks these in standard Pivot Table UI.
Pivot with 3-level hierarchy. SUM(Value), COUNT, AVG(Return), StdDev(Return), Variance(Return). Full distribution in one pass. ~56 seconds.
Survey Response Analysis
20K respondents, 120 question columns (Q1_Satisfaction through Q120_Recommendation). Need long format for R/SPSS. Power Query requires manually selecting each column.
Regex ^Q[0-9]+_.* selects all 120 question columns instantly. Result: 2.4M rows (20K × 120). Save config → reuse for next survey round. 14 seconds.
Under the Hood
Technical capabilities that separate this from Excel add-ins and browser toys.
Pattern Matching for Column Selection (Wildcard + Regex)
Select unpivot columns using * wildcards or full regex — no manual clicking through 50-column lists
Aggregation Memory Model (What's Safe, What's Not)
10 memory-safe aggregations + 2 advanced behind an explicit warning toggle
Hierarchical Pivot (3-Level GROUP BY with Subtotals)
Year → Quarter → Month with drag-drop reordering and subtotal rows per level
Config Save / Load / JSON Export (ConfigManager)
Save 20 configs, export as JSON, share with teammates, import JSON configs — all stored locally
Streaming Web Worker Architecture
50K row chunks, main thread never blocked, pivot O(groups) memory, unpivot O(chunk) flat memory
Perfect For
Not Designed For
Who Uses Pivot & Unpivot
Data Analysts
Transform raw exports into analysis-ready formats. Pivot transaction logs into summary dashboards. Unpivot wide reports into long format for BI tools. No Python setup, no IT ticket, no waiting.
Finance & Operations
Process budget exports, reconciliation files, and transaction logs that exceed Excel's limits. StdDev and Variance for distribution analysis. Recurring report configs for weekly or monthly runs.
Healthcare & Compliance
Process patient records, EHR exports, and claims data without uploading PHI to third-party servers. Client-side architecture is compliant by design. StdDev and Variance for outcome analysis on large cohorts.
How Much Time Are You Losing to Manual Pivoting?
Calculate your annual savings vs. Excel Pivot Tables and Power Query
Typical: 2–8 reports/month
Quarterly = 4, Monthly = 12
Analyst avg: $45–75/hr
Estimates based on a 15-minute manual Excel Pivot Table workflow (setup, field configuration, aggregations, formatting, export). Actual savings vary by report complexity and existing automation. SplitForge processing time varies by hardware, browser, and file structure (±15–20%).
10 Million Rows in ~18 Seconds
548K rows/sec pivot throughput. 680K rows/sec unpivot. Streaming Web Worker — UI never blocks. File never leaves your browser. Also see how SplitForge processes 10M rows in your browser.
Pivot operation: 3 group-by columns, SUM + COUNT + AVG, CSV output
Unpivot operation: 2 ID columns, 48 value columns, CSV output
Variance: Results vary by hardware, browser, aggregation complexity, and file structure (±15–20%)
Frequently Asked Questions
Is my data private?
What file sizes and row counts are supported?
How does pattern matching for unpivot work?
What aggregations are available?
What is hierarchical pivot?
Can I save and reuse configurations?
What output formats are supported?
Does this work with Excel files (.xlsx)?
What are the honest limitations?
Related Guides
Stop Rebuilding the Same Pivot Table Every Monday.
12 aggregations. Pattern-matching unpivot. Save configs for recurring reports. File contents never leave your browser.
Also try: Aggregate & Group · VLOOKUP / Join · Data Profiler · CSV Splitter