DataTerminology.comNEW
Also known as: Data Product Lifecycle Management, DPLM
The stages of a data product from discovery and design through delivery, operation, and retirement.
DESCRIPTION
The data product lifecycle spans ideation, design, build, validation, release, operation, iteration, and sunset. It mirrors software lifecycles with data-specific governance and quality controls. Observability tracks usage, SLAs, and incidents. Versioning and documentation support reproducibility and trust. Automated tests validate schemas, metrics, and policies. Feedback loops guide prioritization and improvements. Retirement plans address archival and compliance. Lifecycle discipline enables sustainable value delivery and reduces technical debt.
ALTERNATIVE DEFINITIONS
Wikipedia
Product lifecycle describes the stages a product goes through from conception through withdrawal from the market.
General lifecycle concept applicable to data products.
MLflow
The Model Registry provides a centralized model store, set of APIs, and UI to manage model lifecycle from experimentation to deployment.
Lifecycle management applied to ML models within data products.
Microsoft Learn
MLOps practices manage the lifecycle of ML models, including training, deployment, monitoring, and retraining.
Operational lifecycle aspects relevant to data product components.
RELATED TERMS
Data Product MLOps Lifecycle Management Model Registry Versioning Observability & Telemetry Governance Retirement Policy
COMPETING TERMS
Ad-hoc Data Delivery
KNOWLEDGE AREAS
MLOps Data Management Data Products Observability & Telemetry Testing & Validation
TARGET AUDIENCE
MLOps Engineer Data Engineer Data Steward Data Architect BI Specialist Analytics Engineer
EXTERNAL SOURCES
- Wikipedia - Product lifecycle
- Microsoft - MLOps production lifecycle
- MLflow - Model lifecycle registry
- Martin Fowler - Data as a product lifecycle context
TOP QUESTIONS
What stages make up the Data Product Lifecycle?
Discovery, design, build, validate, release, operate, iterate, and retire, with governance and quality controls throughout.
Category: Overview
Why is lifecycle management important?
It ensures sustained value, reliability, compliance, and predictable evolution of data products.
Category: Relevancy
How do you measure lifecycle health?
Track adoption, SLA adherence, incidents, schema change success, and versioned documentation completeness.
Category: Measurement
What is a common lifecycle mistake?
Skipping validation and observability, leading to broken dashboards or downstream outages.
Category: Mistakes
How does MLOps support the lifecycle?
It provides pipelines, registries, monitoring, and retraining practices for ML components in data products.
Category: Relation
What risks exist in lifecycle management?
Uncontrolled changes, drift, and lack of deprecation plans can break consumers or violate policies.
Category: Risk
How to begin lifecycle management?
Define SLAs, versioning, testing, and change control; instrument usage and reliability metrics; establish retirement procedures.
Category: HowTo
What competes with disciplined lifecycle management?
Ad-hoc delivery without SLAs or versioning, which undermines trust and reuse.
Category: Alternatives
Who participates in lifecycle management?
Data product managers, engineers, stewards, and platform teams collaborate to maintain quality and compliance.
Category: Collaboration