Building the Data Stack for Operational ML
by:
Mike Del Balso, Co-Founder & CEO, Tecton

Mike will kick off the event and present his views on the different types of features commonly used for Operational ML use cases, and solutions to manage them.

Operational ML models rely on several types of features with different characteristics such as serving latencies, feature freshness, data sources, and transformation pipelines. We’ll categorize the main feature types that are most frequently encountered in real-world use cases, and discuss the specific challenges of building bespoke pipelines for each type.

We’ll present our view on a better approach to managing all common feature types. A feature platform is a system designed to manage the complete lifecycle of features. It decouples feature definitions from feature transformations, providing a higher level of abstraction that simplifies the task of building Operational ML, while using modern data infrastructure and best practice architectural patterns.

apply(): The ML data engineering Conference
Presented by Tecton

Connect with us:
Slack: [ Ссылка ]
LinkedIn: [ Ссылка ]...
Twitter: [ Ссылка ]

свернуть