If you haven't already, please request a free trial to get access to the tutorial.
In this tutorial, we will be using click and impression streaming data to build, evaluate, and safely deploy an ad serving model. Along the way, you will inspect the data, create and register new features to Tecton, and deploy them in production.
This tutorial shows how to use Tecton for easy management of machine learning features in development and production. In order to complete this tutorial, you'll need to use git, the command line, Spark, and python. This short video will give an overview of what you can expect:
The tutorial is comprised of three parts - each designed to get you familiar with Tecton's overall workflow and functionality, and ready to develop on the platform with your own data:
- Build training data and test the serving API in a Notebook. Tecton has sent you a link with a Databricks workspace to use - in your "Welcome to Tecton" email. The "Shared" workspace should have a "Tecton in 10 Minutes" notebook to walk through. See instructions on accessing the tutorial here.
- Navigate the Web UI. Log into your Tecton Web UI - a link for this is also in your "Welcome to Tecton" email. If you'd like a guided walkthrough, you can click on the Navigating the Tecton Web UI tab in this tutorial for a few videos.
- Begin adding new content. Once those are complete, you can set up your own environment and begin creating new features in the Taking the Next Steps tab of this tutorial.
So, let's get started! Please reach out at firstname.lastname@example.org if you need any help along the way!