UiPath Automation Suite

The UiPath Automation Suite Guide

Adding a dedicated agent node for Task Mining

Hardware requirements

Before starting the Task Mining installation, make sure to check Hardware requirements.

After completing installation of the Automation Suite in either single-node evaluation or multi-node HA-ready production mode, run the following steps to set up Task Mining.

Enable Task Mining on the cluster

If you already enabled Task Mining in the configuration during the main install, you can skip this step. If not, you need to enable AI Center and Task Mining by editing the configuration and rerunning the install.

Editing the configuration

The following needs to be added to the configuration file cluster_config.json during installation in the Advanced Configuration step.

aicenter and task_mining should be added as enabled in cluster_config.json.


Task Mining config
    "aicenter": {
        "enabled": Boolean
    "task_mining": {
        "enabled": Boolean

Configuring a dedicated node for Task Mining

Step one: Configure the machine

Follow the steps for configuring a server node to ensure the disk is partitioned correctly.

Step two: Copy the interactive installation wizard to the target machine for installation

For online installation

SSH to the right machine:

  • If you added Task Mining to a single-node evaluation install, go to the machine and run this next step.
  • If you added Task Mining to a multi-node HA-ready production install, go to any of the server nodes.

Run the following command to copy the contents of the UiPathAutomationSuite folder to the Task Task Mining node (username and DNS are specific to the Task Mining node):

sudo su -
scp -r /opt/UiPathAutomationSuite <username>@<node dns>:/opt/
scp -r ~/* <username>@<node dns>:/opt/UiPathAutomationSuite/

For offline installation

SSH to the primary installation node
Ensure that the UiPathAutomationSuite directory contains sf-infra.tar.gz file ( is a part of the installation package download step )

scp -r ~/opt/UiPathAutomationSuite <username>@<node dns>:/var/tmp

Step three: Run the interactive install wizard to configure the dedicated node

For online installation

SSH to the Task Mining Node and run the following commands:

sudo su -
cd /opt/UiPathAutomationSuite
yum install unzip jq -y

UNATTENDED_ACTION="accept_eula,download_bundle,extract_bundle,join_task_mining" ./

For offline installation

Connect via SSH to the Task Mining dedicated node & install the platform bundle on it using the script below.

sudo su 
mv /var/tmp/UiPathAutomationSuite /opt
cd /opt/UiPathAutomationSuite
sudo chmod -R 755 /opt/UiPathAutomationSuite

sudo ./ -i ./cluster_config.json -o ./output.json -k -j task-mining --offline-bundle ./sf-infra.tar.gz --offline-tmp-folder /opt/UiPathAutomationSuite/tmp --install-offline-prereqs --accept-license-agreement

Step four: Enabling kubectl

Run below command to set the kubectl context on taskmining machine

sudo su -
export KUBECONFIG=/var/lib/rancher/rke2/agent/kubelet.kubeconfig
export PATH=$PATH:/var/lib/rancher/rke2/bin
kubectl get nodes

You should see the nodes and their corresponding names. You need the name of the Task Mining node for the next step.

Step five: Verifying task mining configuration

Finally, run the following command to check if the Task Mining node is configured. Please update the <node name> with Task Mining node name in below command.

kubectl describe node <node name> | grep -i "taints"

Above command should return you the following output


At this point, you have successfully completed the installation for Task Mining. Sign in to the Automation Suite Portal and enable it in the tenant management UX under the Admin section.


Configuring email

Be sure to configure system email notifications as Task Mining relies on the Automation Suite notification system to send emails.

Starting a Task Mining project

The steps for creating a task mining project are the same whether you're using the UiPath Automation Cloud or Automation Suite.

Configuring the Task Mining ML Model in AI Center

Follow the steps for the given deployment type, online or offline.

Online and Offline installations

Before proceeding with the steps below, it is mandatory to start with following the steps from Uploading ML Packages.

After uploading the ML package, follow the listed steps:

  1. Access AI Center home page for your tenant and click on the Create Project button. This project can be named anything you want but we recommend matching it to the name of your Task Mining project.

  2. From the project, navigate to the ML Packages page and then choose Out of the box Packages

  1. Select the UiPath Task Mining package.
  1. Select TMAnalyzerModel and click Submit.

:information-source: Please always select the latest version of the model packages.

  1. Name the package, choose the most recent package version, and click Submit to create a machine-learning deployment package.


Increasing secondary disk/Longhorn storage

As Task Mining is a data-heavy application, we recommend an additional 1024 GiB of storage to handle the screenshots and metadata collected during studies. After adding a secondary disk, longhorn storage must be resized to incorporate it.

How to configure a new disk in the cluster

You can add the additional disk either before pre or post installation. If it's added post-installation, you have to perform additional steps to configure this new disk in the cluster.

See Configuring new data disk in the cluster for more details.

How to increase storage capacity

To increase the storage capacity after completing the installation navigate to UiPathAutomationSuite folder and run the below command.

sudo su -

cd /opt/UiPathAutomationSuite

sudo ./ objectstore resize-pvc --size 1000Gi --sub-component data


AI Center dependency

Task Mining has a dependency on AI Center, so AI Center always needs to be installed if Task Mining is installed. This is done as part of the Advanced Configuration step mentioned earlier.

Updated 10 months ago

Adding a dedicated agent node for Task Mining

Suggested Edits are limited on API Reference Pages

You can only suggest edits to Markdown body content, but not to the API spec.