This repository provides AI/ML service(MachineLearning model serving) modernization solution using Amazon SageMaker, AWS CDK, and AWS Serverless services.
Note: This repository worked with Feast 0.9 and before. Please look at http://github.com/feast-dev/feast for the more up to date version of this repo.
Overview
This repository contains the following Feast components.
Feast Core: The central feature registry used to define and manage entities and features.
Feast Serving: A service used to serve the latest feature values to models.
Feast Java SDK: A client used to retrieve features from Feast Serving.
Helm Charts: The repository also contains Helm charts to deploy Feast Core and Feast Serving into a Kubernetes cluster.
Architecture
Feast Core has a dependency on Postgres.
Feast Serving has a dependency on an online store (Redis) for retrieving features. The process of ingesting data into the online store (Redis) is decoupled from the process of reading from it. Please see Feast Spark for more details about ingesting data into the online store.
Feast Serving has a dependency on Feast Core.
The Go and Python Clients are not a part of this repository.
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].
