DataTerminology.comNEW
Versioned outputs from build, training, or analysis processes—such as binaries, packages, models, reports—used for deployment, reproducibility, and auditability.
DESCRIPTION
Artifacts are tangible outputs produced during software and data lifecycles, including compiled binaries, Docker images, model files, and generated documentation. They capture the state of a build or experiment for reuse and verification. Good artifact practices include semantic versioning, immutability, checksums, and metadata like provenance and SBOMs. Artifacts reside in repositories integrated with CI/CD for publishing, promotion, and rollback. In MLOps, artifacts record models, feature sets, and evaluation reports to support experiment tracking.
ALTERNATIVE DEFINITIONS
Wikipedia
In software development, an artifact (or artefact) is one of many kinds of tangible by-products produced during the development of software.
Broadly characterizes artifacts as by-products across the lifecycle.
Microsoft Learn
Artifacts are the files or packages produced by a build.
Concise CI/CD-centric definition tying artifacts to build output.
RELATED TERMS
Build Artifact Model Artifact Data Versioning Artifact Repository Continuous Integration Reproduction Software Supply Chain Checksum Content Addressing Release Candidate
COMPETING TERMS
KNOWLEDGE AREAS
Platform & Tooling Software Engineering MLOps Data Engineering
TARGET AUDIENCE
DevOps Engineer MLOps Engineer Software Developer Analytics Engineer Data Engineer Application Administrator
EXTERNAL SOURCES
- Wikipedia - Artifact (software development)
- MLflow - Artifacts in MLflow tracking
- DVC - Data and model versioning basics
- GitHub - Actions workflow artifacts
TOP QUESTIONS
Why treat Artifacts as immutable?
Immutability guarantees reproducibility and auditability. Changing artifacts undermines trust and complicates debugging and compliance.
Category: Best practices
What metadata should accompany Artifacts?
Include version, checksums, provenance (build info), SBOM, signatures, and dependency graph for security and traceability.
Category: Requirements
How do Artifacts support rollbacks?
Versioned artifacts allow precise rollback to known-good builds or models, reducing mean time to recovery.
Category: Example
Where should Artifacts be stored?
Use an artifact repository or registry that supports your formats, policies, and integrations with CI/CD and deployment tools.
Category: Best practices
What’s the biggest misunderstanding about Artifacts?
Confusing them with source code; artifacts are build outputs, whereas source is the human-edited input to build them.
Category: Misunderstanding
How do Artifacts relate to SBOMs?
SBOMs enumerate artifact components and dependencies, enabling vulnerability scanning and license compliance.
Category: Security
Do data teams produce Artifacts?
Yes—models, feature sets, and reports are artifacts used in deployments and audits.
Category: Relevancy
How to prevent artifact bloat?
Apply retention policies, deduplication, content-addressing, and compress large assets where appropriate.
Category: Mistakes
What is a hot topic around Artifacts?
Provenance standards (SLSA) and universal signing (sigstore) to harden software supply chains.
Category: Topic