Epiq Discover Version 4.7 Release Notes (April 2026)

This release contains updates and improvements to Epiq Discover. The following notes highlight new enhancements to Epiq Discover and provide a list of bugs addressed in this release.

Release Highlights

Release 4.7 contains the following features, as well as additional enhancements listed in this document. The following sections contain details about these features and enhancements.

• Predictive Coding. This release features our new Continuous Active Learning (CAL) based Predictive Coding capability. Predictive Coding learns from Reviewer coding decisions on a small set of documents and applies those patterns to a larger document population.

  After each round of training, the Predictive coding model assigns predictive scores to the entire model population, indicating how likely they are to receive a positive or negative coding value. You can use these scores to build a review queue that prioritizes documents most likely to be coded as positive.

• Epiq Assist. Epiq Assist now streams responses progressively as soon as generation begins.

• Document List. In the Grid view and the Undocked Viewer, you can now filter short text, long text, single choice, multiple choice, boolean, and date columns in the Document List.

• Export updated production files. Added the ability to export updated DAT, image, native, and text files individually or in combination for completed productions without re-exporting the entire production.

Enhancements in Release 4.7

This release contains the following enhancements and updates.

General

Product name update

Epiq has rebranded Epiq Discovery as Epiq Discover and also updated the names of the following products:

• Epiq AI Discovery Assistant is now Epiq AI

• TMX is now Epiq Narrate

Epiq Discover now reflects these updated product names. (1042703, 1042704, 1042705, 1042708, 1042709, 1042710)

Collect

Slack collection update

The Slack collection feature is now enabled by default for all projects and clients. Client Administrators and users with higher roles can enable or disable this feature at the client or project level. (1038014)

Process

ICS file processing update

The system now extracts embedded ICS files contained within a parent ICS file as an individual file for each calendar entry. (1029856)

Inspect/Review

Predictive Coding

This release features our new Continuous Active Learning (CAL) based Predictive Coding capability. Predictive Coding learns from Reviewer coding decisions on a small set of documents and applies those patterns to a larger document population.

After each round of training, the model assigns predictive scores to the entire model population, indicating how likely they are to receive a positive or negative coding value. These scores build a review queue that prioritizes documents most likely to be coded as positive.

Here is the basic workflow for generating and retrieving predictive scores using CAL.

1. Prepare the document population. Create a saved search or move the documents into a folder. Identify the single-choice field and the positive and negative values for coding documents. Then assign these values to some of the documents. Ensure that the document population includes at least one document coded as positive and one coded as negative in order to run the initial training of the model. Adding more positively and negatively coded examples improves the model’s accuracy.

2. Create a predictive model. When creating the model, specify the document population using the saved search or folder you created, and select the field along with the positive and negative values you identified earlier.

3. Create an assignment for the predictive model you created. The system generates review batches that include the highest-priority documents identified by the predictive model. You can retrieve batches from the prioritized queue after the model is initially trained. The batches include only the promoted documents.

4. Review batches. Reviewers review each retrieved batch and code documents with a positive or negative value. For any coded documents in completed or released batches, reviewer coding decisions are used to retrain the model for more accurate predictions. The system sends only the positive and negative coding decisions to the model for retraining. The retrained model generates the predictive scores for the entire document set and updates the review queue according to the predictions.

5. Retrieve the predictive scores. When the model generates the predictive scores, you can retrieve them and view them in the field labeled <model name> Score. The scores are searchable and exportable for analysis and reporting purposes. At any point in the process, you can retrieve the model’s predictive scores to review the score range for the retrieved batches and to assess the accuracy of the predictions against reviewer coding decisions.

Also, you can add more documents to an existing Predictive Coding model using Document Actions>AI Model and retrieve their predictive scores. When you add documents using this option, the system transfers the coding values associated with the documents.

Also, the AI Model field displays Chat when used for chat functionality, Predictive Coding when used for predictive coding, or both values when used for both features.

To view your model metrics and assess the maturity of the model, access Epiq AI. Refer to Epiq Suite shortcuts for more information.

The Predictive Coding feature has the following limitations.

• For best performance at this time, we recommend limiting uploads to no more than 250K documents. While you can upload larger sets, Epiq AI divides them into a set of smaller uploads per model with no more than 2 GB or a maximum of 250K files. In addition, it is recommended to limit the number of models to less than 100.

• The system excludes any documents that do not contain extracted text.

The following list provides related topics.

• Create a predictive model

• Create an assignment

• View an assignment

• Review a document

• Retrieve predictive scores

• View Create Predictive Model job details

• Add documents to AI models in Inspect

• Add documents to AI models

• View AI Model job details

(918728, 1012552, 1013919, 1013969, 1013976, 1018232, 1018804, 1019191, 1019996, 1021819, 1021831, 1021832, 1024165, 1025136, 1025561, 1012547, 1018796, 1030075, 1024389)

Epiq Assist updates

Feature name update

In the Main Menu, updated the Chat name to Epiq Assist. (1044520)

Added the ability to stream partial responses

Epiq Assist now streams responses progressively as soon as generation begins. This improves responsiveness and reduces perceived latency. When response generation begins, the Stop icon temporarily replaces the Submit icon in the prompt box. You can select Stop at any time to stop response generation. (1000554)

• While generating a response, Epiq Assist displays a loading indicator.

• When you stop response generation, Epiq Assist displays the partial response and adds Stopped Generating at the bottom.

• If you click Stop before any response is generated, Epiq Assist displays Stopped Generating.

• During response generation, you can perform the following actions.

  • Type text in the input box. However, you must stop response generation before sending your prompt.

  • Select a chat from History. The system saves the partial response.

  • Select New Chat to immediately clear the conversation.

Document List: Added the ability to apply filters

In the Grid view and the Undocked Viewer, you can now filter short text, long text, single choice, multiple choice, boolean, and date columns in the Document List.

To apply filters, in a column header, click Filter , and then do the following.

• For short text, long text, single choice, multiple choice, and Boolean, select one or more values.

• For date fields, choose an operator and select a date or date range.

• Select Is Null to return documents with no values in the field.

• Click Apply.

When a column has an active filter, the Filter icon remains visible. At the top of the Document List, Filter By shows the applied filters.

To remove filters, do one of the following.

• To remove a filter from a single column, click Filter and then Reset.

• To remove all filters, click next to Filter By.

Also, in the Grid view, you can now sort the Document List as you do in the Undocked Viewer Document List.

(1038479, 1041166, 1038480, 1038478, 1038478)

Production export update

Export updated production files

Added the ability to export updated DAT, image, native, and text files individually or in combination for completed productions without re-exporting the entire production. In Document Actions>Export, when you select Export Production in the Select Export Action field, the system now displays the Output Format and Production Template fields. In the Output Format field, you can select one or more options such as Data, Images, Natives, or Text and export only those files in a zipped package, without exporting the entire production again.

In the Production Template field, the system displays the Production Template name associated with the selected production. Click the Edit icon to edit template settings as needed. In the Production Template, only the sections that match the output formats you selected are enabled for editing, while the others remain disabled. If no output formats are selected, opening the template shows all sections as disabled. After selecting your output formats and updating the template as needed, the system exports only the selected file types and packages them for download. (871979)

Added Missing Document Report

When exporting a production, if any of the selected documents are not included in the chosen production, the system now generates a Missing Document Report, which lists the missing documents.

To download the report, in the Jobs Overview page or the JOBS page in Inspect or Review, click Download for the export job. (1004980)

Export to PDF update

Added a new option named Original Images Setting in the Export to PDF document action, allowing you to select whether original images are exported using the original format, as searchable PDFs, or as non‑searchable PDFs. This option is only visible when you select Output Selection as Original Images. By default, the system selects Use Original Format, and you can change this setting to Make PDFs Searchable or Make PDFs Non‑searchable as needed. The system applies the selected output type for both individual and merged PDF exports and displays the selected option in the Export to PDF job details pop-up.

When you select Natives as Output Selection, the system checks for native files only. If native files are not available, the system displays the affected documents in the error report with the message The selected documents do not have native files. The same behavior applies when you select Original Images as the Output Selection. (1004295, 1024488, 1032219, 1034475, 1032219)

Also, the system now successfully completes the Export to PDF of produced images without errors or file overwrites, naming each PDF with the beginning Bates number and organizing them into folders by production volume. (1032219)

Persistent redaction settings update

In Document List>IMAGE Viewer, when you select redaction options such as fill type, fill color, or background color, the system now retains these settings as your default setting until the session ends. (772822)

Mid‑level archive file promotion and export update

For projects created in Epiq Discover release 4.7 or later, the system now sets the Document State field of mid‑level archive files such as email archives and compressed or packaged formats to ExcludedArchive during processing when extraction is successful. As a result, the system excludes these files from promotion and export. (892081)

• If the system does not successfully extract mid‑level archive files, it does not set the Document State field to ExcludedArchive. Therefore, the system can promote or export these files.

• The system does not process OneNote (.one) files as Excluded Archive.

Produce

Production update

In Production Template>TEXT, updated the name of the Set Precedence field to Export Precedence. Additionally, the OCR Text and Extracted Text options in this field are now preselected. For existing templates, the original text precedence configuration is retained without changes. (1019504)

Administration

Report update

Added a new report named Project Access Report. This report displays User Name, User Role, Project Name, Login, Logout, and the Duration (Minutes) between those times. The system records the login time when you open a project and the logout time when you leave the project for any reason, using a timestamp formatted as yyyy‑mm‑dd‑hh‑mm‑ss‑ms in 24‑hour format with milliseconds. You can find the new report in the following areas.

• Account Menu > Client Reports > Report Type

• Projects Jobs page

• Client Report list

To create a Project Access Report, complete the following steps.

1. In the Account Menu, select Client Reports.
2. In the Client Report dialog box, in Report Type, select the Project Access Report.
3. In Users, click to select all users or select specific users.
4. In Date Range, choose a Start and End date.
5. In Projects, click to select all projects or select specific projects.
6. Click Create.
7. You can view the Project Access Report in the Projects Jobs page or Client Report list.

Also, updated the User Login History Report. This report no longer displays the Action column. Instead, it now shows User Name, User Role, Login, Logout, and Duration (Minutes), which captures the time between login and logout. (1023147, 1030772, 1052793)

Desktop Client

Session timeout update

The Desktop Client session timeout is now configurable. The default timeout is 30 minutes, and you can configure it to any value from 30 minutes to 4 hours. To change the session timeout, contact your Epiq Discover Administrator. (1038231)

Bugs fixed in Release 4.7

The following bugs were resolved in this version.

Number	Title
847783	Viewer | Native view is not able to display the document and shows the error message "Remote end closed connection without response."
989208	Processing | Responsive text missing in native view of HTML Document
989509	Modify Metadata | CustodianID Incorrectly updated as Name in OS due to Metadata job failure
1001632	Transfer to Relativity | Parent Fragment enters infinite loop when the heartbeat is not sent properly to the assigned task ID
1010588	Undocked viewer | Document load delay in Undocked Viewer – 7-8 Seconds per document
1026311	Transfer to Relativity | Transfer report shows errors despite successful document load into Relativity due to error "System.Runtime.CompilerServices.ConfiguredTaskAwaitable`1. ConfiguredTaskAwaiter.GetResult() at Relativity.Services.Proxy.RPC.KeplerRPCClient."
1028502	Promotion | Promotion with "Within Set" deduplication results in inconsistent duplicate handling
1029083	React | Coding Form | Duplicate lower case field value created
1031284	Client Dashboard | Archived projects shows 0 for hosted size (current and peak)
1034641	Client Dashboard | Incorrect display of the current hosted size and peak hosted size for archived projects
1034894	Transfer to Relativity | Documents are not transferred and the job shows completed due to OutOfMemoryException
1036236	Epiq Internal Feedback | Manage Fields | 6 default fields that are already under the "Selected" column can be selected twice
1036329	Loading | Attachments are not getting promoted because of missing metadata
1036369	Conceal | Multiple final fragments are created for the Conceal job
1036441	When you remove auto‑redactions, the Audit History Report now correctly shows the metadata values in the Old Value and New Value columns. The report no longer shows Deleted as a metadata value for any field, including Redaction Type or Has Redactions. If a document contains only auto‑redactions and you remove them, the system updates the Has Redactions value from Yes to blank
1036915	Transfer to Relativity | Unauthorized error during transfer execution
1037150	Saved Search | Duplicate save attempts and long save duration
1039191	Corrupted (?) .msg file attachments causing extraction failures
1045453	Epiq Internal Feedback | Clients reporting slow loading times in native viewer when reviewing
1049164	Search Builder | Previously built search conditions with user created fields are not pulled back for edit
1049458	Processing | Processing job completes with errors "java.lang.StackOverflowError"
1051022	Search Builder | Incorrect search results in older projects when using multiple tags with “includes any of these "option in Search Builder
1051590	AppProxy is not timing out application correctly