Uploading Files to Harvey

Review Tables are designed to support up to 10k documents. However, you may experience slower performance, during loading or while running the table, when working with a very large number of columns and files. We’re actively improving performance to better support larger tables, as well as increase our file limits.

Addressing Delays or Timeouts for Review Table Uploads

If you notice performance issues, consider splitting your data into multiple Review Tables. You can do this by:

• Dividing files across separate tables (e.g., split 100 files into two tables of 50 each), or
• Reducing columns per table (e.g., keep the same 100 files, but split into separate tables with fewer columns each)

If your file exceeds the size limit, here are some steps you can take:

1. Compress the file – Zip the file to reduce file size. For instructions, perform a search in your web browser for, “how to zip a file in [your browser].”
2. Upload files in smaller batches or individually.
3. Use alternative formats – Some formats (like PDF or JPEG) are smaller than others while keeping quality.
4. Check for unnecessary content – Remove extra pages, images, or data that aren’t needed.

Why Uploads Can Take Time in Vault

During a Vault upload, Harvey also processes those files for search and use in the platform. This processing takes more time than a standard file transfer.

How to Address Any Upload Delay Across Harvey

If your upload is slow, stuck, or timing out, try the steps below.

Quick Fixes

1. Wait 5 minutes and retry the upload.

   1. Temporary system or browser slowdowns often resolve on their own.
2. Disconnect from your VPN, if connected.

   1. VPNs can slow large uploads or cause timeouts. Disconnect briefly and try again.
3. Switch networks if possible.

   1. For example, switch from office Wi-Fi to a hotspot or home network.

Optimize Large Uploads

1. Upload in Smaller Batches. Instead of uploading everything at once:

   1. Start with ~1,000 files
   2. If successful, increase to 2,000
   3. Then scale up gradually
   4. For best performance, keep batches under 10,000 files
   5. Why this helps:

      1. Prevents browser crashes
      2. Reduces the risk of losing progress
      3. Helps you identify your system’s limits
2. Split large single files into smaller files to upload. For example:

   1. Excel: Break worksheets into multiple files, or segment the data by rows when possible. If formulas reference other worksheets, keep those worksheets in the same file so Harvey can trace calculations across them.
   2. PDF: Use Adobe Acrobat’s PDF splitter to separate a large file into multiple files. Keep each PDF under 10,000 pages.
3. Upload From Local Storage: Files stored on network drives or cloud-synced folders upload more slowly due to added data transfer overhead (network-stored files require your local machine to manage both network communication and upload processing).

   1. For better performance, download files locally before uploading.
   2. If available, use a direct integration supported by Harvey (e.g., iManage) instead of uploading manually from a network drive.
4. Upload files to one vault at a time rather than uploading to multiple vaults in parallel.
5. Ensure your firm’s corporate security tools support Harvey: Some security tools (e.g., firewall rules) block large file uploads or restrict access to certain cloud services. If uploads fail repeatedly, contact your IT team and let them know the platform may need to be allowlisted.

If none of the above resolves the issue, proceed to General Troubleshooting.

Uploads fail if the file type is unsupported, or if the file extension (e.g., .pdf) does not match the actual content type (e.g., Word document).

1. Confirm your files are within Harvey's Supported File Types and Size Limits listed earlier in this article.
2. Check that file names and formats match.

   1. For example, a file named NDA.pdf that’s actually a Word document, should be named NDA.docx.
3. Try opening the file, save it as a new file, or convert it to a PDF before re-uploading.