Your files, in one workspace.
Browse, inspect and work with your files. Keep useful results together, then save a portable workspace or download a copy.
Files and tasks stay in this tab. Nothing is uploaded.
File tools Assemble PDFs, clean and compare data
Try a sample workflow
More examples
Find across your files
Search open PDFs, spreadsheets, SQLite tables and supported members of open folders or ZIPs. Open a result at its page, row, cell or member.
Search the files already open in this tab.
Clean → Compare
Apply saved cleanup recipes to fresh before/after CSV exports, match IDs, and export the differences. Keep a reusable job without storing records.
Choose the earlier export, or use two CSVs already open in the workspace.
Choose the newer export.
Reusable workflow job
Download a settings-only job after a successful comparison. Next time, load that job and choose fresh CSV files. Recipe filters, output fields, keys and mappings are saved; source rows are not.
No saved workflow loaded. Choose a cleanup recipe for each side, or leave either side unchanged.
Choose two CSV files. This task stays in the current tab when you navigate.
Comparison settings
Adjust these fields, then recompare without reopening the files.Fields to compare (0 mapped)
Choose which field from B should be compared with each A field. Unmapped fields are intentionally excluded.
Comparison report
Everything runs locally in this tab. Originals are untouched. An invalid new file, incompatible recipe or changed schema preserves the last prepared report rather than silently remapping fields. Jobs contain settings only, and are never stored automatically.
Split a CSV into grouped copies
Choose a category column and create a separate CSV for each exact group. Review group counts before preparing a ZIP; every source row is kept.
Categories match exactly, including letter case and surrounding spaces. Blank or whitespace-only categories form a separate missing group. Up to 100 groups; no rows are deduplicated or filtered.
Choose a CSV file and column to split into groups.
Review group counts
ZIP filenames are numbered group-001.csv, group-002.csv etc., rather than exposing categories in paths. Manifest.csv maps each file to its exact category and row count. The manifest and files contain source values; review before sharing. All processing remains local. Inputs and prepared results stay in this tab until you reset or discard the task, or close the tab. Downloads are explicit.
Summarize a CSV by group
Count records by a category, or total and average a numeric column. Export a compact summary without changing or uploading the source.
Groups use exact source values (case and spaces matter); empty or whitespace-only categories share a distinct missing bucket. Amounts must be dot-decimal numbers with up to six decimal places. Missing amounts are counted separately; invalid numbers stop the summary rather than silently becoming zero.
Choose a CSV file to summarize locally.
Review the grouped totals
Read-only and browser-local. At most 25,000 records, 32 columns, 250,000 cells and 2,000 groups; summary output is capped at 2 MB. Exact arithmetic avoids floating-point drift; averages round half away from zero to six decimal places. Changing options clears the old result. Switching views keeps your task in this tab; reset or discard it to release the data. Only explicit downloads create files.
Match two CSV exports by ID
Join rows side by side using exact identifiers, without guessing about duplicate keys or merging empty IDs. This is different from Combine CSV, which appends rows vertically.
A · Main export
B · Additional export
Choose two CSV files and their matching identifier columns.
Review matches before exporting
Duplicate nonempty IDs stop the operation instead of pairing arbitrary rows. Empty IDs remain separate. Comparison is exact and case-sensitive, preserving leading zeros and whitespace.
Read-only and local. No files are uploaded or automatically saved. The files and results stay in this tab until you reset or discard the task, or close the tab. At most 20,000 data records and 32 fields per source, 250,000 output cells and a 10 MB export.
Combine multiple CSV exports
Append records from 2–10 CSV files into one table. Match headings by name even when columns move, preview differences, and explicitly download a new copy. This does not join or deduplicate records.
Records are appended in the order you select files, preserving their order within each file. Header names are trimmed when matching, but case-only disagreements, repeated names, empty headers and uneven record widths are rejected.
Choose 2–10 CSV exports to inspect their headers.
Review column alignment
Union must be selected explicitly. No source record is dropped or deduplicated. Formula-like cells are escaped for safe spreadsheet opening.
Everything stays in your browser. Files are read-only; switching views keeps this task in the sidebar until you reset or discard it. Nothing is uploaded, persisted, or downloaded until you choose an export action. Maximum combined output: 50,000 records, 250,000 cells and 10 MB.
Check a CSV before cleaning it
Find missing values, repeated full rows, repeated identifier values, ambiguous headings and uneven records. Review a local report without changing the file.
Identifier values are matched exactly, including spaces and letter case. Repeated values are findings to review, not automatically errors. A quoted multiline field counts as one CSV record.
Choose a CSV file to check it locally.
Fields
Headers and record widths
Repeated rows and identifiers
The downloaded report has no scripts or external resources. It may include sample identifier values, so review before sharing. No uploads or automatic persistence. Inputs and results stay in this tab when you switch views. Reset or discard the task to release them; reloading or closing the tab also clears them.
Repeat a cleanup across files
Use a previously downloaded Data Tools recipe. Process 2–10 CSV files locally, then export one ZIP of cleaned copies. Originals are never modified.
1 Choose a saved Data Tools recipe
Choose a saved recipe from Data Tools.
In Data Tools, expand “Save or reuse these settings” and choose “Download recipe”.2 Choose CSV exports
Choose 2–10 CSV exports, up to 2 MB each and 10 MB total.
Choose a Data Tools recipe and 2–10 CSV files.
All-or-nothing: if any file is missing required columns or has no matching rows, no new ZIP is prepared. The first row of each CSV provides column names. Input files stay in this tab only and are released when you reset or discard Batch Clean, or close the tab. No uploads or automatic saving.
Compare datasets
Choose your before and after files, or use data already open. Match records by key and download the differences.
Repeat a saved comparison
Save the sheets, keys and field mappings for next time. Jobs contain settings and field names; choose fresh input files for each run.
Run a comparison to save its setup, or load a previously downloaded job.
Choose two open datasets.
Choose a list, keep the fields you need, then copy or open it as a table. Your original file stays unchanged.
Use commas and ranges (3-1 reverses three pages). Duplicates are allowed; maximum 200 output pages. Move or remove pages below. The source stays unchanged. Interactive forms and encrypted PDFs are unsupported; bookmarks and signatures are not preserved.
Enter the PDF password.