II.
Workflow overview
Reference · liveworkflow:feature-store-management
Feature Store Management overview
Maintains the shared feature store — registering new features, monitoring freshness SLAs, decommissioning stale features, and ensuring consistent feature computation between training and serving. Excludes feature engineering research.
Attributes
displayName
Feature Store Management
workflowKind
data
triggerType
continuous
typicalCadence
continuous
complexity
single-team
description
Maintains the shared feature store — registering new features, monitoring
freshness SLAs, decommissioning stale features, and ensuring consistent
feature computation between training and serving. Excludes feature engineering research.
Outgoing edges
applies_to_domain2
- domain:data-science·DomainData Science
- domain:ml-ops·DomainMLOps
involves_role3
- role:ml-engineer·RoleMachine Learning Engineer
- role:data-scientist·RoleData Scientist
- role:platform-engineer·RolePlatform Engineer
performed_by_org_unit2
- org-unit:ml-platform-team·OrgUnitML Platform Team
- org-unit:data-platform-team·OrgUnitData Platform Team
requires_skill_area2
- skill-area:python-data-pipelines·SkillAreaPython Data Pipelines
- skill-area:kafka-stream-processing·SkillAreaKafka Stream Processing
triggers_responsibility2
- responsibility:data-quality-monitoring·ResponsibilityData quality monitoring
- responsibility:slo-definition·ResponsibilitySLO definition
Incoming edges
follows_workflow1
- stack-profile:feature-store-mlops·StackProfileFeature Store & MLOps Stack (Feast, MLflow, BentoML, K8s, Prometheus)