Example code for feature store best practices
This project has been created for demonstrating code to implement a Feature Store.
The following assumptions were made for implementing the demonstration under sample-code:
- System requirements: Snowflake and SageMaker
- Feature Store technology preferences: none
- Security: Feature Store access will be kept behind a company firewall on a private VPC; no requirement for exposure
- Scale: Petabytes scaling towards exabyte level data lake
- Budget and performance constraints: typical
- Feature rate of change: typical
- Edge deployments: Assuming not required short term, but longer term may be desirable
The directory contains the following files:
.
├── LICENSE: Permissive license
├── README.md: This file
└── sample-code: directory of sample code demonstrating best practices for each of AWS Feature Store and HopsWorks, as comments indicate.
├── 1-Centralization.py
├── 2-Version_control.py
├── 3-Scalability_and_performance.py
├── 4-Data_quality_and_monitoring.py
├── 5-Security_and_access_control.py
├── 6-Integration_with_ML_workflow.py
└── 7-Documentation_and_metadata_management.py