Skip to content

Introducing Tecton's Framework v2

We're excited to introduce you to Tecton's Framework v2! While the core concepts remain the same, we've improved the API and added popular new features based on user feedback.

This section of the documentation contains the updated Overviews and How-to Guides for using the new API.

We'll be working with all our current customers to transition to this new framework soon. In the meantime, you're welcome to test out the new framework, and please reach out if you have any questions.

Getting started

Framework v2 is available for you to start testing today! In order to do so, you just need to update your Tecton SDK:

  • Update your CLI
  • Update the Tecton SDK for your Notebook Cluster on Databricks by simply restarting the cluster, or cloning the cluster on EMR

Once you're up to date, you can create new Tecton Objects in your feature repository.

What's Framework v2?

  1. Transformation definitions now have
    • Streamlined authoring for single-transformation features; and
    • Flexible pipelines for composition and re-use.
  2. On Demand Feature Views (formerly Online Feature Packages) can now combine current request data with materialized batch or stream features.
  3. BatchFeatureViews formerly TemporalFeaturePackage) can join multiple batch sources in their transformations.
  4. Backfilling for Batch Feaure Views (formerly TemporalFeaturePackage) with data look back will be much more efficient.
  5. New object names make it easier to differentiate batch and stream processing.

New Tecton Object Names

Framework Concept Diagram

This diagram illustrates how Tecton Objects interact with each other in the new framework.

While we're changing some terminology, the primary concepts will be familiar to any existing Tecton users.

At a high level:

  • BatchDataSource and StreamDataSource replace VirtualDataSource for defining external data sources
  • FeatureTable replaces PushFeaturePackage for defining features transformed outside of Tecton
  • BatchFeatureView and StreamFeatureView replace TemporalFeaturePackage, while BatchWindowAggregateFeatureView and StreamWindowAggregateFeatureView replace TemporalAggregateFeaturePackage.
  • Transformations are now more modular, and single-transformation Feature Views are easier to define.
  • OnDemandFeatureView replaces OnlineFeaturePackage for defining features based on request-time data.