eDiscovery AI Installation
INSTALLATION AND CONFIGURATION GUIDE
1. Compatibility
The eDiscovery AI Relativity application is compatible with the following Relativity versions:
· Relativity One
· Relativity Server 2022 – 12.1.537.3 Patch 3
· Relativity Server 2023 – 12.3.857.3
2. Installation – eDiscovery AI Relativity application
Go to the Application Library in Relativity.
Only System Administrator users can perform the installation.
In the Application Library, click on the Upload Application button.
The system opens the Application Information page, click on the Select File button.
Click on “Select File” and Browse for the Application Rap file or drag and drop the RAP file on the section.
2.1Relativity RAPs
Based in the version of Relativity must be select the RAP.
· Relativity One the RAP is eDiscovery AI_Relativtiy.rap,
· Relativity Server the RAP is Server_eDiscovery AI.Relativity.rap
The system will upload the Rap file to the Relativity instance. Once the application information is loaded, click the Save button.
The installation process of the application will start in the Relativity instance. Wait for the installation process to be completed in the Relativity instance.
2.2 Workspace Installation
In the section “Workspaces Installed” Click on “Select” button to install the application in some workspaces.
The system will open a modal with the list of workspaces, select the workspaces where the application needs to be installed and click on the "Apply" button.
Wait until the application process is completed on the selected workspaces.
In the event of a failure during the installation in the workspace, an error message will be displayed in the status.
To address this issue, follow these steps:
1. Click on the error message to open a modal that contains detailed error information.
2. If the error message indicates "Name Conflict: Extracted Text Size," follow these steps to resolve it:
a. Choose the "Map Field" option.
b. Click on "Retry" to attempt the installation again.
Upon completing these actions, the installation process should be successfully completed.
3. Agents
During the installation process, the following Agents will be automatically created
in the instance. Each Agent is responsible for different automated
processes.
- Upload Planner: Fetches one request at a time per execution.
- Upload Worker: Fetches 100 documents at a time per execution.
- Request Worker: Fetches one request at a time per execution.
- Retrieve Worker: Fetches one request at a time per execution.
- Propagation Worker: Fetches one request at a time per execution.
The application requires the above-mentioned agents to be added and active as a minimum for it to function as expected.
3.1 Agents Creation
It is possible to create new Agents in the following cases.
Case1: If any of these agents are missing which is possible in the OnePrem versions.
Case2: It is possible to add more agents to the environment to increase availability and/or improve the application's performance.
Please follow the following steps to create them.
1. Go to the "Agents" tab and click on the "New Agent" button.
2. Click on the "Select" button next to "Agent Type" and choose the agent you want to create. You can filter the agents by typing "eDiscovery" in the application name field.
3. Once the agent is selected, click on "Set."
4. The number of agents needed will depend on the usage of the application, as each one collects a different amount of data:
- Upload Planner: Fetches one request at a time per execution.
- Upload Worker: Fetches 100 documents at a time per execution.
- Request Worker: Fetches one request at a time per execution.
- Retrieve Worker: Fetches one request at a time per execution.
- Propagation Worker: Fetches one request at a time per execution.
Note: Please be mindful that each agent consumes memory, so having a large number of them is not recommended unless you have the infrastructure to support it.
4. Click on the "Select" button next to "Agent Server" and choose the server that best suits your needs. Once the server is selected, click on "Set".
5. It is recommended to set the Run Interval field to 5.
6. Finally, click on "Save" to create the agent, and repeat the process as many times as necessary for the required number of agents.
By following these steps, you will have the required agents set up to ensure smooth operations within the application.
4. Instance Settings
In the installation process, the following instance settings will be automatically created in the instance as part of the application:
· ApiKey
o Name: ApiKey
o Section: eDiscoveryAI.Connect
o Encrypted: Yes
o Description: ApiKey to connect to eDiscovery AI API
· SizeLimitationInKB
o Name: SizeLimitationInKB
o Section: eDiscoveryAI.Connect
o Encrypted: No
o Default Value: 1000
o Description: Size limitation to uploading files to the eDiscoveryAI API
· eDiscoveryAIEndpoint
o Section: eDiscoveryAI.Connect
o Name: eDiscoveryAIEndpoint
o Value: api.ediscoveryai.com
o Encrypted: No
o Description: The endpoint of the eDiscoveryAI API
5. Installation Set long text field size application
The eDiscovery AI Relativity application utilizes the "Extracted Text Size" column to calculate the file size.
It is necessary to configure this field in the instance, use the "Set Long Text Field Size" application and follow the installation process outlined for the eDiscovery AI Relativity application.
6. New issues in the Request
The eDiscovery AI Relativity application offers 15 available issues to add to each request by default; however, only 5 are visible in the "View Request."
To adjust the number of issues, you can edit the layout and include the new issues in the view. This action can be performed by users with permission to edit layouts or by administrators.
6.1 Requests - Data Sent to eDiscovery AI API
All requests generated in Relativity will include the following parameters:
· Instance_name: Name of the Relativity instance from which the job is sent.
· Workspace_id: Id of the workspace from which the job is sent.
· Workspace_name: Name of the workspace from which the job is sent.
· Issues: the list of the issues the user setup in the mass operation layout
7. Initial Access Permission
7.1 Prior to your first document submission, you will need to provide
your eDiscovery AI representative with the originating Public source IP
address(es) of your Relativity Server instance. For security purposes,
eDiscovery AI restricts communications from all but pre-approved sources.
7.2 If you have direct access to your Relativity Agent Server, you may
check this with a web browser by going to www.ipdonkey.com
7.3 If you are unable to access your Relativity servers or if they are
not permitted browse the internet, please consult your Network/IT Support to
determine the correct public IP address. The IP address we require is not the
locally assigned or private address (typically beginning with 192.168 , 172, or
10) but rather the address that is used by a network router or firewall to
access the internet at large.
8. Troubleshooting communication failure
8.1 Upon document submission, if you see:
8.2 Your Relativity Agent server(s) may be restricted from the
internet in some way from a firewall or other security system.
Please consult with your Network or IT
security team to request that this server have access to the following:
US / Global
· https://api.ediscoveryai.com
· https://ediscoveryai.blob.core.windows.net
UK
• https://api-uk.ediscoveryai.com
• https://ediscoveryaiuk.blob.core.windows.net
Related Articles
eDiscovery AI Prerequisites
eDiscovery AI Prerequisites Before conducting any type of eDiscovery AI review, the following prerequisites must be met. 1) System Requirements a) Extracted Text - All text based (i.e. emails, word docs) documents must have extracted populated in the ...Using eDiscovery AI in Relativity
Using eDiscovery AI in Relativity 1) eDiscovery AI fields and objects can be added to your Relativity Layouts and Views like any other object in Relativity. b) To add any of the eDiscovery AI fields to your layout, scroll or filter until you’ve found ...eDiscovery AI Job Tracking & Metrics
1) Tracking and Job Metrics a) After the documents have been submitted to eDiscovery AI you can monitor the progress of active jobs and view statistics across all jobs. b) Tracking Progress i) Monitor the progress of the AI analysis in real-time ...