Driving Efficiency and Quality through a Feature Store at Delivery Hero

Wednesday, October 11, 2:15pm - 2:45pm (EDT)

Time shown in-04:00 America, New York

Speaker: Breno Costa

At Delivery Hero, cross-functional teams within the Data & ML product line are committed to building a great personalized experience for our customers. They achieve that goal by working on different kinds of ML models including ranking and recommender systems, consumer segmentation, incentive science, and so on.
To support those efforts and improve the long-term efficiency of our teams, the ML platform team has built a Feature Store that empowers data scientists to efficiently create feature transformations in SQL or Python, monitor features using quality checks and drift detection, and serve offline and online features for their machine learning models.
Our feature store solution has established a standard feature engineering process. Using a standard framework, data scientists follow good practices for the development and validation of features. With a shared feature repository, data scientists leverage existing features, increasing feature reusability across projects and reducing duplication of similar features. Automated feature pipelines reduce engineering efforts for application teams to maintain pipelines in production.

Add to Calendar 2023/10/11 11:15:00 2023/10/11 11:45:00 America/Los_Angeles Driving Efficiency and Quality through a Feature Store at Delivery Hero Speaker: Breno Costa

At Delivery Hero, cross-functional teams within the Data & ML product line are committed to building a great personalized experience for our customers. They achieve that goal by working on different kinds of ML models including ranking and recommender systems, consumer segmentation, incentive science, and so on.
To support those efforts and improve the long-term efficiency of our teams, the ML platform team has built a Feature Store that empowers data scientists to efficiently create feature transformations in SQL or Python, monitor features using quality checks and drift detection, and serve offline and online features for their machine learning models.
Our feature store solution has established a standard feature engineering process. Using a standard framework, data scientists follow good practices for the development and validation of features. With a shared feature repository, data scientists leverage existing features, increasing feature reusability across projects and reducing duplication of similar features. Automated feature pipelines reduce engineering efforts for application teams to maintain pipelines in production. false MM/DD/YYYY 60 OPAQUE apQfLtmRnzOiFbgoNmal132273

Link: