Initial Exploration

Persona: Data Scientist (primary). Also relevant: AI Engineer.
In this module

Get hands-on with the platform: create an OpenShift AI Workbench, connect to the repo, and open your first notebook so you’re ready for later labs.

Estimated time: 20–30 minutes

What you’ll do

  • Use Console Links to navigate

  • Create a GPU-enabled Workbench

  • Clone the lab repo into your Workbench

  • Open and run the getting started notebook as a data scientist

Getting Started

Explore the environment

Use the OpenShift Console Links to quickly navigate around the environment.

  • RHOAI

  • LLamaStack Playground

  • OpenShift

  • ACM

  • Argo CD

Console Links

Create a workbench

  1. Login to OpenShift AI and select the agent-demo data science project.

    Create Workbench

  2. We are going to Create a workbench using the following parameters:

    Name: agent-tools
Image Selection: Standard Data Science
Version: 2025.1 (select the latest version)

    Leave all the other fields as defaults. You should see the Hardware profile auto-selected to use the GPU Accelerator Nvidida L4 (Shared).

    CUDA Workbench

    Select Create workbench.

  3. Once the workbench is running open it in your browser.

Open the first notebook in your workbench

  1. Clone the code into your workbench by using the Terminal and entering:

    git clone https://github.com/redhat-ai-services/etx-agentic-ai.git
    Clone Code

  2. Open up the following notebook in your workspace.

    etx-agentic-ai/code/getting-started.ipynb

    The getting-started.ipynb notebook will be empty at this stage. We’re including it here to demonstrate how a data scientist would typically interact with the repository and set up their environment. In later modules, you’ll use this notebook as a starting point for building and testing your own agents.