Glossary
A guide to terms used throughout the Baseten platform and documentation.
These terms appear throughout the Baseten platform and documentation. We strive to use industry-standard language, but sometimes multiple standards exist or we legitimately need a new word to describe a new thing. Terms are presented in alphabetical order.
A full-stack web application and API composed of any number of views, worklets, and files that can call any number of models and data tables.
A hosted and managed database within your Baseten account, available to queries and worklets.
A reference to a worklet or query result, component value, state, query parameter, or other piece of information within a view.
An individual step within a worklet that performs a piece of data processing, machine learning, or business logic.
A connection to a data source outside of Baseten (MySQL, PostgreSQL, BigQuery, Snowflake, Redis, Kafka, SageMaker, S3).
A single table within Baseten Postgres or a data connection that is available to queries and worklets.
The overall structure of a data table, including the name and type of each column.
A trained machine learning model deployed on Baseten and available to applications.
A set of pre-trained models that solve common machine learning tasks.
A saved SQL query written against Baseten Postgres or a data connection, used as a source of data for worklets and views.
A web interface for interacting with an application deployed on Baseten.
The drag-and-drop WYSIWYG editor for creating views and wiring them up to queries and worklets.
A serverless function that surrounds a model invocation with business logic and glue code.
Last modified 7mo ago