Skip to navigation
User GuideExamples and Notebooks

Notebooks

View as Markdown

Download this quickstart notebook to learn how Kumo Relational turns connected business tables into predictions without training a task-specific model. The example uses four tables from Microsoft’s AdventureWorks sample database.

In the notebook, you will:

  • Install and configure the NVIDIA Kumo Relational client.
  • Connect to a deployed Kumo Relational NVIDIA NIM inference service.
  • Load and prepare the sample data as pandas DataFrames.
  • Build and inspect a relational graph, which maps the tables and their key-based relationships.
  • Submit prediction requests for several use cases.

Before you begin, you need a deployed Kumo Relational NIM endpoint and the AdventureWorks data files identified in the notebook.