Skip to main content
Version: 0.5



Before you begin, we recommend reading the Tecton Concepts page for an understanding of key Tecton concepts and product interfaces.

Follow these steps to set up your environment prior to beginning the tutorial.

Connect Tecton to a data platform (if not already done so)

If your Tecton instance is not already connected to a data platform (Databricks or EMR), follow these steps:


  1. Follow the steps in Configuring Databricks.

  2. Follow the steps in Connecting Databricks Notebooks.


  1. Follow the steps in Configuring EMR.

  2. Follow the steps in Connecting EMR Notebooks up to, but not including, the section Configure the notebook.

Install the Tecton CLI

If you have not already done so, install the Tecton CLI on your local machine.

Clone the tutorial feature repository

Clone the tutorial feature repository, which is located on Github.

This repository contains the skeleton structure for the feature pipelines you will build, and is incomplete. As you work through this tutorial, you will add code for the features and other feature-related code.

Log in to your Tecton account with the Tecton CLI

In your terminal, run:

tecton login <account-name>

Replace <account-name> with your Tecton account name. This is the same as the URL for your Web UI.

Logging into your account will allow you to create and manage workspaces, apply feature repo definitions to a workspace, manage API keys, and more.

Create a workspace

You will initially create a development workspace to use for testing your features during early development. Later, you will create a live workspace to productionize features.

Run the following command to create a development workspace, replacing <workspace name> with a name of your choosing, such as tecton-fundamentals-tutorial.

tecton workspace create <workspace name>

Create a new notebook

Create a new notebook in Databricks or EMR.

Install JARs in your notebook (EMR only)

If using EMR, configure the notebook by installing the Tecton JARs.

Import necessary modules in your notebook

In your notebook, run:

%pip install mlflow

Then, run the following in the next notebook cell:

import tecton
import pandas as pd
from datetime import datetime, timedelta

Retrieve a reference to your new workspace

In your next notebook cell, run:

ws = tecton.get_workspace("<workspace name>")

This workspace object (ws) will be used to read feature definitions from the workspace.

Was this page helpful?

Happy React is loading...