Upload Native Data to Everlaw

Everlaw has a cloud-processing system that processes and ingests native files into the platform. Documents processed on Everlaw go through de-NISTing, deduping, OCR, AV transcription, and language detection, as appropriate. Additionally, Everlaw generates text, metadata, and PDFs (if requested) for your native data.

A native upload can contain one file, or multiple files.

• For more information about preparing your files for upload, please see our article on Preparing Native Data for Upload
• For information about managing completed uploads, see our article on how to Manage Native Uploads

You can use this article to:

• Guide you step-by-step through a native upload 
• Learn about managing native uploads once they are on Everlaw

To read more about uploading documents on Everlaw, please refer to the articles in our Uploads section.  

Requirements

To access the upload page: You must be a Database Administrator, have Upload permissions on the database, or be an Organization Administrator.

To upload into any specific project: You must be either a Project Admin or have Partial Project Document Management for that project

Step 1: Start the upload

To access the Uploads page and start the upload:

1. From the homepage, go to Data Transfer > Uploads.
2. Select Native, then Start upload.
   
    
3.  You can upload files from your computer, or from a cloud connection to an app or cloud-based storage service. 
   
   To upload files locally from you computer, select Browse to select files from anywhere on your computer, or drag and drop them in. Please see our article on Preparing Native Data for Upload for information on how to prepare your files prior to upload. 
   The following section on uploading via cloud-based apps provides details on uploading via cloud connection.
4. When you're done, select Next to move onto the Details step. 

Upload via cloud-based apps

To upload via app, select an app to begin uploading your documents.  

Once you select an app, you are asked to log in via a separate dialog box. If you do not see this dialog box, check your pop-up settings. After logging in, you are able to select your files directly from the app's platform. Then, you move on to the Dataset Details step. 
For all uploads completed via cloud connector, any top-level folder uploaded via cloud storage source will be compressed and displayed in Everlaw as a container file.

Select an app name to open an article with more information about it:

• Purview
• Asana
• Microsoft 365 (also known as Office 365, O365, or M365)
• Jira
• Slack
• Zoom
• Zendesk
• Salesforce

You can also upload via cloud based storage apps: 

• Box
• Dropbox
• Google Drive 
• SharePoint
• OneDrive
• ShareFile
• Google Vault (You can also upload exported Google Vault files)

Upload via direct link 

Once you’ve made your selection, for files, you are taken to the Dataset details step.

Step 2: Details

In this step, you specify the configuration for your upload. By default, your upload settings are inherited from the last upload in this database. 

To complete the Details: 

1. In the Name field, enter a unique name for your upload.   
2. Choose your Native data deduplication setting. Select one of the following:
   • Global: Deduplicate against all of the existing documents that have been natively uploaded in your database.
   • By custodian: Deduplicate against all of the existing natively uploaded documents by custodian. This means that if two duplicates have different custodians, they will both be uploaded. Conversely, if a document has the same custodian as another duplicate document that already exists on the platform, the duplicate file will not be uploaded.

   • None: No deduplication.To learn more about the definition of duplicates and how upload deduplication handles documents (and families), visit this article on duplicates.

     Note

     Google files (e.g., Google Documents, Google Sheets) will not get deduplicated in the same way as other file types, because they undergo their own conversion process within Google.

3. If, in the Native data deduplication step, you selected either By custodian or Global, you also have the option to select Deduplicate these documents against processed documents in the database. Select this option if you want to deduplicate documents in your current native upload against processed documents already in the database. You might do this if:

   • You have migrated native data from a different platform to Everlaw, and uploaded the migrated data as processed data or via our migration process. You want to deduplicate this native upload against this data.
   • You have reopened an Everlaw database by uploading the previously exported documents as processed data, and want to deduplicate this native data against the restored data
   • You have already uploaded received productions as processed data, and think that some data in your native upload might be duplicative of this data

   Here are some important details about this option:

   • The option you selected above for native deduplication (By custodian or Global) will apply to deduplication against processed data as well

   • The processed documents you deduplicate against must include a native file and the SHA1 hash value of the native document and native file of the processed document must match. 

     Note

     Everlaw uses the hash value of the native file. It does not use the hash metadata value included in the load file

   • Deduplication against processed data can result in new documents within your current native upload being deduplicated out. If a parent/container file is a duplicate of a processed document already in the database, the native container/parent along with all children/attachments will be deduplicated out of the upload, even if the attachments are not duplicates of anything already in the database.
   • If there are multiple duplicates that are eligible to be deduplicated against (e.g. both a natively uploaded document and a processed document), the “primary duplicate” that is used to deduplicate against is always the document that was uploaded first

   Note

   To deduplicate against processed data uploaded on or prior to January 14, 2026 (January 13 in Australia), you must reprocess the processed data before you complete the native upload.

4. If any of your files/folders are password-protected, enter the password(s) into the Passwords for protected files field (one password per line) to enable Everlaw to image and extract text based on your processing options. Password-protected files for which you do not enter a password will not be processed. The native view is not available for password-protected documents on Everlaw.
5. [Optional] Select the caret under any of the sections under Advanced settings to configure settings for Language and location, Image details, Chat documents, and Decryption. The section below on Advanced settings includes details about each option:

• Language and location
• OCR Language
• Image Details
• Chat documents
• Decryption Keys
• DeNIST settings
  For more details, see the Native data processing settings section.

6. When you're done, select Next to move onto the Custodians step.

Advanced settings

Under Advanced settings there are several ways to configure your upload.  This section describes the options for each advanced setting.

Language and location

OCR Language

This setting specifies particular languages for Optical Character Recognition (OCR).  By default, OCR language is set to Autodetect. Autodetect can extract all Latin-alphabet languages (such as French and German) as well as Chinese, Japanese, and Korean (CJK). If your document has a combination of these aforementioned languages, Autodetect will also be able to OCR them automatically as long as there is only one language per page. Autodetect will not reliably OCR multiple languages within one page.
 

OCR is automatically run on all image files (TIFFs, JPGs, PNGs, etc.), inlined images in emails, and PDF pages with fewer than 50 embedded non-whitespace characters. Handwriting on these documents can be detected and OCRed, but the OCR tool used during upload is not optimized for handwriting and the quality of the output depends on the quality and legibility of the handwriting. After upload, you can reprocess documents an force OCR using a tool optimized to capture handwriting in English.

Here are some additional tips about selecting an OCR language:

• Select Autodetect if your upload includes multiple documents, each in different languages. 
  However, this means that non-Latin, non-CJK language documents will not get properly OCRed. For example, if one document is entirely in Arabic (non-latin, non-CJK language), and another is in French (Latin language), then only the French document will be properly OCRed.
  In this situation, you can either separate those documents into different uploads and select the appropriate OCR language for each or, after upload, select specific subsets of documents from the results page for reprocessing with a different OCR language.
• Select a single language to target for OCR if there is a specific non-Latin and non-CJK language in the upload. OCR will only detect that language and English. There are two scenarios where you would want to select something other than Autodetect:
  • If all your documents are not in a Latin-alphabet language or CJK (e.g. Russian, Greek).
    In this case, you must select that language from the dropdown menu in order for OCR to work for that document.
  • If the quality of your scanned document(s) is low, and you know there is only one language in the document (in addition to English).
    This improves the quality of OCR, but only for that language and English. It will prevent the detection of any other languages in that document, so you should be sure that the document only has only one non-English language before selecting that option.
  • You can also use the OCR language field to specify transcription of Spanish files with extractable audio. Select Spanish from the OCR language dropdown. You cannot transcribe Spanish files and OCR other documents with non-English languages in one upload. You can always reprocess Spanish media files and transcribe them in Spanish once they’re uploaded.

Default Timezone

The selected timezone is assumed for any datetime metadata lacking an explicit timezone. Metadata without an explicit timezone is given the selected timezone on Everlaw.
Email headers printed at the top of PDF images generated during processing show datetimes in the selected timezone.

If you do not want to use a default timezone, select the X in the entry field.

Image Details

Expand the caret next to Image details to update the settings for Create PDFs for, Page size, Inline images in attachments, Hyperlinked images, and PowerPoint speaker notes.
Create PDFs for

By default, Everlaw creates PDF images for all files in an upload, and placeholder images for file types that don't image well, such as spreadsheets. You can instead choose: 

• All documents and deselect Do not image spreadsheets and large text files: This option will image the file types that don’t image well. In addition to Excel files, the following file types are not be imaged by default, and, but are imaged when you choose this configuration: 
  • LibreOffice Calc
  • Empty files
  • Container files
  • iWorkNumbers
  • QuattroPro
  • TXT files that are greater than 5 MB
• No documents: No images are created for any of the files in an upload
  

Page Size

Everlaw generates PDFs in the selected size for documents that do not have a described size. Documents with an explicit size (e.g. PDFs, word documents, and images) will remain in their original sizes. 
Emails do not typically have a described size, but if they do, they will remain in their original sizes on the platform. Documents exported, printed or produced from Everlaw will respect the size of the pages on the platform. 

Email image attachments

There are three options for deciding whether image attachments should be displayed inline, or treated as separate attachments:

• Inline all images found in emails: Every image in the email is displayed inline within the PDF
• Extract all images found in emails: Email image attachments are extracted as children of the parent email. Images that are explicitly inlined (e.g. pasted into the email body) will not be extracted . 

• Smart determination: Everlaw will dynamically determine which images are likely to be attachments, and which ones are (or are intended to be) inlined images (e.g., signature icons). Factors influencing this smart determination include an image's dimensions, overall size, and content ID. 
   

  Tip

  Since Smart determination makes dynamic decisions, this option typically leads to a more common-sense set of extracted attachments. 

Hyperlinked images

Choose how to process images that are hyperlinked in an email.:

• Fetch hyperlinked images: Everlaw attempts to fetch linked images that appeared in the body of the email to the original recipient. The image presented is the image located at the URL at the time of processing, which is potentially different from the image at the time the email was sent.
  If the original (native) email contains a link to an image, but the image itself was not displayed  when the email was originally sent, the image will not be fetched and displayed by Everlaw upon processing. Instead, the link will be present in the body of the text, just as in the original email.
• Do not fetch hyperlinked images: Images referenced by URL are not fetched or displayed

PowerPoint speaker notes

Choose how any included PowerPoint speaker notes are formatted:

• Include speaker notes with original formatting: Speaker notes of .pptx, .ppt, and .pps files are included in the PDF file with their original formatting—the "Notes Page" view in the PowerPoint application. Notes with overflow text will continue onto the following page. 
• Include speaker notes with optimized formatting: With this option, the notes scale with the slides, the page orientation is landscape, and the speaker notes text are displayed under the slide. Each PDF page contains the corresponding slide number, and overflow note text is continued on the following page. Any customization of the “Notes Page” view in PowerPoint, such as the addition of headers and footers, is not reflected in the PDF or text files.
• Exclude speaker notes from PDF and text file: The slide is full page in the PDF file with no notes present in the PDF or text file. Speaker notes are still included in the native view and searchable via the Speaker Notes metadata search term.

Chat documents

If your upload includes chat data, you can specify additional processing details:

Segmentation

Choose how the chat messages will be divided across documents. You can choose to Split conversations by number of messages or Split conversations into daily segments (24 hour period).
If you select Split conversations by number of messages, you also select a segment size:

• Small segments: Each document contains a maximum of 25 messages
• Medium segments (default): Each document contains a maximum of 50 messages
• Large segments: Each document contains a maximum of 100 messages

If you Split chat conversations into daily segments, each chat segment will contain messages sent in a single 24-hour period in the upload time zone.

Here are a few additional notes about the segmentation options:

• Once the data is loaded on Everlaw, each chat document has no more than the maximum number of chat messages selected on this step.
• Many chat platforms offer threading. When segmenting chat conversations by number of messages, Everlaw will generally keep threaded messages together. If a thread is long enough that the maximum size of the segment is exceeded, threads will be split across documents.
• You cannot reprocess chat documents to update their segmentation

Slack attachments

Everlaw can fetch Slack attachment files uploaded into Slack, like images, PDFs, Word documents, and media files. Including attachments with your Slack upload can result in a data size increase of your uploaded chats. You can choose:

• Fetch attachments referenced in Slack messages to include these attachments
• Do not fetch attachments referenced in Slack messages if you do not want to include them in your upload.

The option you select applies to both local uploads and those through the Slack cloud connector. When you select not to fetch attachments during a local upload, any attachments included in the ZIP folder will not be extracted.
If you do not include attachments upon upload and want to include them later, you are able to reprocess your Slack upload to include them.

Decryption Keys

Everlaw can store private keys used to decrypt S/MIME encrypted emails. To learn more, read this article. You can follow this link to manage your decryption keys on a new tab.

DeNIST settings

DeNISTing is the process in which system files and other non-user-generated data are removed from a collection of ESI.

In this section, you can choose between:

• DeNIST (default): Excludes all files from the NIST list
• Do no DeNIST: Includes files from the NIST list

Step 3: Custodians

In the Custodians step, you specify what custodian value to associate with the documents you’re uploading. You can specify a default custodian for all documents in an upload and/or set custom custodian values for particular files or folders. If your data belongs to multiple custodians, please read How to Handle Uploads With Multiple Custodians to learn how to prepare your data accordingly before uploading.

To set the custodian(s):

1. [Optional] Input the custodian name into the Default custodian box at the top of the table. If your project already has custodians from legal holds or previous uploads, you can also select one from the dropdown list.
2. To set custom custodian values for particular files or folders, find the file/folder on the table. You can either input the custodian name into the custodian box or use the pencil icon next to each file/folder on the table to autofill the custodians using file/folder name
   Alternatively, select the Autofill all visible custodians to autofill all custodian values in the table based on visible file/folder names.
   
   Here are additional tips for assigning the proper custodian values: 
   • Files that have a black caret in the far left can be expanded to display the individual sub-folders/files they contain. Click on the caret icon to expand or collapse. The custodian value for non-visible files (i.e. files that are hidden because the parent folders are collapsed) is assigned based on the closest visible parent. 
   • If there is a default custodian, it will be overridden for that particular file/folder with the custom value.
   • You can use the pencil icon next to each file/folder on the table to autofill the custodians using file/folder name.
   • Select Clear all to delete custodian values for all files, regardless of whether they are visible in the table. If there is a default custodian added, all custodian values will be changed to the default value.
     
      
3. When you're done, select Next to move on to the Projects step.

Step 4: Projects

In addition to uploading the documents into the current project you’re on, you can add the documents to partial projects for which you have Partial Project Document Management permission on. Here are details about uploading into projects:

• Documents are always added to all complete projects within the database
• Documents are always added to the project from which you have accessed the uploader
• There are no billing implications for adding data to multiple projects within the database

When you deduplicate your upload and then choose to upload documents to a partial project, it's possible that a document in your upload is a duplicate of an existing document in your database, but that existing document is not yet part of the partial project. When this happens, the existing document is added to the partial project.

To help you better identify such documents, Everlaw creates a new binder called "Deduplicated docs from dataset [Name]" in the affected partial project. This binder contains documents that (1) were not part of the partial project prior to the upload and (2) are now part of the partial project because they are existing versions of deduplicated documents from the upload. 

To select the additional project(s) to add these documents to:

1. Select the checkbox next to the project name. Deselect a checkbox if you do not want to upload documents into that project. 
   You cannot deselect complete projects.
2. When you're done, select Next to move on to the Additional options step.

Step 6: Additional options

On the Additional options step,  you can apply work product (ratings, codes, binders, notes, etc.) and custom metadata values to documents in your upload. For details and instructions, see our article Applying Custom Metadata Values and Work Product During Upload.

Step 7: Summary

The Summary step provides an overview of your upload, including configuration settings, projects being uploaded into, applied work product, and applied custom values. 

When you have confirmed that the details look good, select Upload. Otherwise, select Previous to make edits.

Transfer and Processing

Once you select Upload, your data begins transferring to Everlaw. The Upload details dialog appears on the Transferring tab and shows you the status of the transfer.

From the dialog, you can add additional documents to the upload by selecting + Add files.

A card is added to the Native Data Uploads page corresponding to your upload. Once processing starts, a time estimate appears on the status card to indicate approximately how long it will be until processing is complete. 

As processing continues, you can start reviewing completed documents. Once all your files are successfully processed, the status card has a green checkmark next to the number of documents in the upload. Select the number to go to a results table of your documents.  

Native uploads are each assigned a control number, indicated by a # prefix.

To learn how to view upload status, delete, rename, and take other actions, please see the Manage native uploads article.

For a comprehensive guide on troubleshooting errors, please see this article on native upload errors.

Native data processing settings

Here are a few additional details about native processing on Everlaw:

• The orientation of documents is preserved from its native version (e.g., a document that is in landscape orientation will remain that way upon upload)
• Embedded files: Everlaw will extract all embedded files, including audio and video files, in a Microsoft 365 file (e.g., an Excel file embedded in a PowerPoint or a PDF embedded in an Excel file) and any file embedded in a PDF. These files are identified on Everlaw as attachments (children) to the document from which they were extracted.
• Emails containing URL links to other supported documents will be recognized. See the context panel article to learn more.
• The children of container files are extracted with no limitation on depth. For example, a Word document attached to an email that’s attached to another email that’s in a Zip file that’s in another Zip file will be extracted.
• Hidden columns in Excel are displayed.
• Notes are extracted and presented in the PDF/Image view for Word documents and the Native view for spreadsheets.

• Sometimes, you may try to upload an entire hard drive or a folder with personal files mixed in with system/software files. Some of these files have no user-specific content and can be removed upon processing. This process is called deNIST (removing NIST files). Any files that are on the NIST list, including empty files, qualify for deNISTing automatically upon upload. Binary files, and virtually all containers, are not part of this list and will not be removed.

  Note

  Everlaw ignores the following file types: Windows shortcuts, __MACOSX files, and Thumbs.db. If these files were included in the upload, there is no record of them in the upload report or in Everlaw.