Short message investigative workflow

You can search and code messages in the Short Message Viewer using the investigative workflow. This workflow makes searching and reviewing RSMF faster and more precise, especially when you need to search the short message metadata. For example, this workflow can be used to search messages based on date or time, messages sent by specific participants, edited or deleted messages, and any messages with reactions. Any messages that you find which are relevant to your case can be coded to make it easy to find later.

To search short message metadata and messages, you can use Elasticsearch. To learn more, see Short message search index. Once you are done searching, you can code at the message-level in the Short Message Viewer. This makes finding relevant messages in a document faster. To learn more, visit Short message coding.

Before you begin

Permissions

The following permission is needed to use the Short message workflow:

Object Security
  • Document
    • Message-Level Coding

Installing the short message search application

Before using short message search, install the Search AI application in a workspace:

  1. Navigate to the Relativity Applications tab.
  2. Click New Relativity Application.
  3. Click on Select from Application Library.
  4. Click the Ellipsis icon in the Choose from Application Library field.
  5. Select Search AI.
    This list only displays applications added to the Application Library. If Search AI is not included in the list, see Installing applications.
  6. Click Ok.
  7. Click Import.
    The Search AI app is installed in the workspace and you can now build an Elasticsearch index.

Considerations

This workflow only works with RSMF files processed in Relativity. Any existing RSMF files in the workspace processed before July 14, 2025 need to be reprocessed. To use this workflow, any RSMF files published before July 14, 2025 need to be republished and the Elasticsearch index needs to be built after publishing the files.

  • The workflow features including search and message-level coding are initially configured for specific workspaces. These features are not available in workspaces that were not configured by Relativity.
  • The Elasticsearch index is selectable within search indexes. Since Elasticsearch follows a different syntax, you cannot run searches using the search bar. Short message search is only available via the condition cards.
  • Elasticsearch only works with RSMF files. You do not need to configure the index.
  • Message-level metadata is stored in the cloud and cannot currently be mapped to fields. The RSMF document-level metadata remains unchanged and can be mapped to fields.
  • The fields and choices for message coding cannot be configured and are not propagated to the document.
  • Sliced RSMFs cannot be coded or searched unless reprocessed.
  • Current analysis tools like Communications Analysis operate based on the headers within the RSMF file and not on the messages. Visualizations are based on what is in the header and not the frequency of messages sent.

Short message investigative workflow

You can import, search, and review short message data in RelativityOne, using the following workflow:

  1. You can use Relativity’s Collect for direct access and to automatically convert chat data to an RSMF file. To learn more, see Short message data collection.
  2. If you need to convert .json exports from Slack or .pst exports from Microsoft Teams, RelativityOne Slack and Teams conversion can be configured in your processing profile to convert RSMF files for you. Relativity also provides developer tools to complete RSMF file creation. You may find partners on the Relativity App Hub that also offer RSMF creation services. To learn more, see Importing and converting to RSMF.
  3. You can also take advantage of the advanced searching capabilities using short message search to search across messages and message-level metadata to further cull down data or find additional insights. To learn more, see Searching messages.
  4. Using the Short Message Viewer, review and code the documents using the message-level coding feature. To learn more, see Reviewing RSMF documents.

Short message data collection

Collection

Using the proper credentials, configure and collect data from Slack, Microsoft Teams, Google Chat, or Bloomberg using the Collect application. To learn more, see Collect. To learn more about collecting from Slack, see Short message conversion for Slack.

Data is automatically converted to RSMF and should be converted into 24 hour files by conversation. Consider breaking up extremely dense conversations that contain more than 10,000 messages further into 12 or even 8 hour segments.

Importing and converting to RSMF

If you have received Slack .json and Microsoft Teams .pst exports, you can convert these files to the RSMF format using Processing. Use import to bring in the files and leverage the conversion functionality in Processing.

For other short message data types, use a third-party solution to convert exports to an RSMF. To find a third-party solution, see Relativity App Hub.

Create RSMF files by conversation or channel in 24 hour increments.

Processing an RSMF file

Processing is the recommended method of ingesting RSMF files into Relativity. Using Processing ensures that the appropriate metadata header fields are extracted and families and attachments within the RSMF file are properly linked to give you the best near-native review experience. Processing also ensures message-level metadata is available for short message search and message-level coding.

The metadata contains all the message, timestamps, participants, and reactions from the RSMF file. You can verify that a file should have metadata collected by navigating to the Files tab in processing. Documents with metadata will have a document_uuid attribute. If a document does not contain a document_uuid, you can create a new processing set to process the document and capture the short message metadata.

When documents are published using processing, you can create a document_uuid field and then map the document_uuid field so that it is visible on the Documents tab after publishing. To learn more, see Mapping the document_uuid field.

Before you begin

We recommend doing the following before you process RSMF files:

  • Ensure that the default processing profile has the Extract children option set to Yes. Disabling this field results in missing attachments.
  • Ensure that you select a relational field such as Group Identifier for the Source field. Setting up the Group Identifier in this manner ensures that attachments appear in the Short Message Viewer and are tied to the correct parent messages.

The Advanced Settings section

  • Ensure that the processing profile is setup for Microsoft Teams and Slack.

The Short Message Conversion Settings section

Processing

Once you have made the preparations above, you are ready to process the RSMF files. To learn more, see Processing an RSMF file.

Integration Points

You can now promote RSMF files between workspaces through Integration Points and export the message-level metadata without having to re-process the files. To learn more, see Promoting data between workspaces through Integration Points. The message-level metadata and message-level coding is transferred between workspaces.

If the RSMF files were processed before July 14, 2025, the files need to be republished. To learn more, see Publishing files.

If data was published before July 14, 2025, you will need to republish the RSMF files before initiating the Integration Point workflow. To republish RSMF files:

  1. Navigate to the Files tab.
  2. Select the documents that need to be republished from the list.
  3. Select the Republish mass operation.
  4. Click Republish.

Searching messages

Once processing is complete, you can cull the documents further before starting the review process.

Short message search is available for message-level searching but it does not support search term reports. You can leverage dtSearch for search term reports before reviewing documents.

Short message search returns RSMF documents that have hits and then you can open each document from the Documents tab to see the full conversation. To learn more, visit Short message search index.

Reviewing RSMF documents

Once non-relevant messages have been culled, you can review documents.

Coding RSMF documents in the Short Message Viewer

You can now code individual messages using the Short Message Viewer. To learn more, visit Short message coding.

Coding messages

You can review messages and code them individually as Responsive or Not Responsive, Privileged or Not Privileged and you can add Notes. Editing the layout to add or remove these fields or choices is not currently supported.

The Short Message Coding Pane.

Coding messages and not the RSMF document causes Review Center to not recognize a document as coded.

Performing QC on RSMF documents

After you’ve finished coding messages, you can use short message search to identify all RSMF documents that contain messages coded as Responsive, Privileged, or with Notes. Message coding is stored in the cloud and can only be searched short message search. You do not need to rebuild your index after coding messages. To learn how to identify coded messages in the searchable fields such as Responsive, Privileged, and Notes, see Short message search.

Troubleshooting

If an RSMF document will not build in the Elasticsearch index or you cannot code messages, the message-level metadata may not have been properly extracted from the RSMF. You can do the following to troubleshoot:

  1. If there are no document_uuid values after processing RSMF files, try reprocessing them.
  2. If there are no document_uuid values on the Documents tab after publishing RSMF files, try republishing them.
  3. If you are using Integration Points to export RSMFs to a new workspace and there are no document_uuid values in the destination workspace but there are document_uuid values in the source workspace, check for Integration Points errors and run Integration Points again if there was an error.
  4. If you are using ARM and there are no document_uuid values in the destination workspace but there are document_uuid values in the source workspace, check the ARM settings, including processing, and re-run ARM.

Mapping the document_uuid field

To create and map the document_uuid field:

  1. Navigate to the Fields tab.
  2. Click New Field.
  3. Enter or select the following options in the Field Information section:
    • Name-enter document_uuid using your keyboard.
    • Object Type-Ensure that Document is selected.
    • Field Type-Select Fixed-Length Text from the drop-down menu.
  4. In the Field Settings section, edit the Maximum Length value to 36.
  5. Click on the Advanced Settings tab.
  6. On the Source field, click Select.
    The Select Item - Source modal displays.
  7. Select the document_uuid field from the list and click Set.
  8. Click Save.
    The document_uuid field is created and mapped.
    Once the document_uuid is created, it can be added to a view or saved search so that you can troubleshoot RSMF files. To learn more, see Views and Creating or editing a saved search.