Welcome to the Tecton blog
Our thoughts on extracting ML signal from the data noise
apply() Highlight: How Feature Logging Enables Real-Time ML
Generating training data for an online, real-time machine learning system can be tricky. In order to guard against temporal data leakage, training events must use only historical features that were valid as of that point in time. This requires effectively … Read More
Real-Time Aggregation Features for Machine Learning (Part 2)
In the following sections, we describe an approach to solving these challenges that we’ve proven out at scale at Tecton, and that has been used successfully in production at Airbnb and Uber for several years. The discussed approach explains Tecton’s implementation which relies entirely on open source technologies … Read More
Real-Time Aggregation Features for Machine Learning (Part 1)
Machine Learning features are derived from an organization’s raw data and provide a signal to an ML model. A very common type of feature transformation is a rolling time window aggregation. For example, you may use the rolling 30-minute order count of a restaurant to predict the order preparation time of your favorite food delivery service … Read More
apply() Conference Recap
Last month, on April 21 and 22, Tecton hosted apply(): the ML data engineering conference to share the latest best practices in ML data engineering. The conference brought together industry thought leaders and practitioners from over 30 leading organizations and … Read More
Tecton Named Gartner Cool Vendor
We’re excited to announce that Tecton has been named a 2021 Gartner Cool Vendor in Enterprise AI Operationalization and Engineering(1). This is great recognition for Tecton’s vision of solving the data problem for ML by making feature stores accessible to … Read More
Announcing Feast 0.10
Today, we’re announcing Feast 0.10, an important milestone towards our vision for a lightweight feature store. Feast is an open source feature store that helps you serve features in production. It prevents feature leakage by building training datasets from your … Read More
How to Build a Fraud Model with a Feature Store
Many companies with platforms that involve financial transactions are looking to bring them in-house to some degree as they can have more control over the user experience and usually save money on transaction fees over outsourced solutions. While the advantages … Read More
Announcing apply()’s Speaker Lineup
Tecton is hosting apply(): the ML Data Engineering conference on April 21 and 22, bringing together industry thought leaders and practitioners from over 30 organizations to share and discuss the current and future state of ML data engineering. This conference … Read More
How Machine Learning Teams Share and Reuse Features
What does an “ML-enabled” company look like? The companies that come to mind, like Uber, Twitter, or Google, have tens of thousands of machine learning (ML) models in production. They use these models to make intelligent predictions across the business … Read More