From your systems. With your context.
Bring the full picture.
Not just a sample.
Prepare a complete dataset across the tools your business uses. Start with the source, add the context, and review exactly what you intend to share.
Already started?
Return to a saved submission.
Read your account’s server records, then reselect original files if a transfer needs to continue. File bytes are never saved in browser storage.
Opening this record replaces the local workspace in this tab. Unsaved context and file selections will be cleared. It does not upload or delete anything.
Start with the source
Where does your
data come from?
Select every source you plan to include. We’ll tailor the export checklist. These are file-based submissions, not live connections.
Have a mixed dataset? Choose multiple sources. You can assign each file to its source when you build the inventory.
Make the export useful
Keep the context.
Check the boundaries.
Use your organization’s approved export process. Availability depends on your product plan, admin permissions, and retention settings.
We do not sign into your tools, fetch private links, or request credentials. Prepare the files yourself before selecting them here.
For every source
- Choose the authorized scope and date range. Include complete exports for that scope, not only example records.
- Keep stable record IDs, timestamps, relationships, and a field dictionary where available.
- Review attachments and nested archives. Remove passwords, API keys, session tokens, and unnecessary sensitive information before selecting files.
- Record omissions and access limits. “Full dataset” does not override permissions or include systems you are not authorized to share.
Give the data a frame
A little context.
A more useful dataset.
High-level descriptions only—do not paste actual records, personal details, or confidential content into these fields. Required fields are marked *.
The complete inventory
Your files.
Still on your device.
Select the complete exports you’re authorized to share. Add multiple files or a folder. File selection reads names, sizes, and basic file metadata—not the contents.
Add your dataset files
Drag files here, or choose them from your device.
Archives and individual files are supported.
Upload capacity is checked with the server before a transfer. You can still build an inventory locally.
Selected on this device
Large file sets are shown 50 at a time. Your downloaded manifest includes every selected file. Relative folder paths are included in the reviewed inventory and upload metadata. Use an archive when the internal structure must be preserved together.
No automatic redaction, anonymization, or archive inspection takes place here. Review the actual contents in your own approved tools before uploading.
Permission before transfer
Set the boundaries.
Before you share.
A full dataset can contain many different rights and sensitivity levels. Confirm your authority and disclose the relevant limits.
Do not upload regulated or otherwise restricted records until your organization has approved the destination and handling arrangements. If the contents are unknown, review them first.
One last, deliberate look
Ready for
your review.
Check the metadata and file inventory below. Nothing has been sent.
File inventory
Download your dataset manifest.
A JSON file of your context, source checklist, and complete inventory. It contains file names and metadata, not file contents. Keep it somewhere appropriate.
Selected files remain on your device. Your local manifest is ready to download; no submission or transfer has been made.
Signing in may reload this page and clear this local draft. Download the manifest first. Saved server records can be reopened, but local files must be reselected.
Sign in with ChatGPTThis is a server-side draft, not an upload. This local workspace still clears when you close or refresh the tab.
Keep this tab open during the upload. After a reload, use “Find my submissions” and reselect the original files; file bytes are not saved in browser storage.
A deliberate restart
Your saved context.
Your original files.
Review the saved context and permissions
Reselect the same unchanged original files to continue an incomplete transfer. Matching uses name, size, modification time, source, and relative folder path—not a cryptographic content check. Verified complete files do not need to be reselected.
Incomplete uploads generally expire after about seven days. An interrupted server operation may remain locked and require canceling the submission and starting again.
Recorded file inventory
Choose the same source and use folder selection if the record includes relative paths. Ambiguous, changed, or unrecognized files are rejected; nothing is sent while matching.
Add files that were never registered
A server record cannot include files that were only selected locally before a reload. Add those separately. Files that overlap an existing name, source, and path must be reselected above, not added again.
Dataset transfer
A clear view of
what’s arrived.
Progress below reflects bytes confirmed by the server. It is not a claim that the dataset has been reviewed or accepted commercially.
0 B confirmed
This stops the transfer and asks the server to delete files already received and discard incomplete uploads. Metadata and inventory tombstones remain. Deletion is not complete until the server confirms cleanup; your original device files are unchanged.
Every selected file was confirmed complete and the dataset was finalized. This confirms receipt only—not a review, licensing agreement, valuation, or purchase.
Clarity, before collection
The important
details.
Does selecting a file upload it?
No. Selection creates an in-memory inventory in this tab using file names, sizes, types, and modification dates. Contents are not parsed locally or sent until you explicitly confirm an enabled upload.
Can I include the full dataset?
Yes—prepare the entire scope you are authorized to share, including multiple exports and archives. Configured file, dataset, and file-count limits are shown in the inventory step. Larger datasets need an agreed intake arrangement; this page does not bypass server limits.
Do you connect to Slack, email, or our database?
No. This intake accepts manually prepared file exports when uploads are enabled. Source selection tailors the preparation checklist; it does not establish a live connection or request account credentials.
What happens to personal or confidential data?
This tool does not automatically remove it. Review and minimize the contents with your own approved tools, disclose sensitivity, and confirm appropriate authority and handling arrangements before any transfer.