MLOps and DevOps Terminology and Pipeline Tasks from Literature Review and Interviews

Monika Steidl from Universität Innsbruck compiled a dataset from a multivocal literature review of 151 formal and informal sources and nine semi-structured interviews. The dataset provides a foundation for comparing terminologies for DevOps, CI/CD for AI, MLOps, and lifecycle management. It also includes potential triggers for pipeline reiteration and consolidated tasks for a continuous end-to-end lifecycle pipeline.

Use Cases

• Compare terminologies for DevOps, CI/CD for AI, and MLOps based on the extracted literature review.
• Identify triggers for reiterating an AI lifecycle pipeline based on the consolidated interview data.
• Structure pipeline tasks into Data, Model, Dev, and Ops stages based on the categorized task list.
• Analyze challenges in AI lifecycle pipelines based on the documented information.

Strengths

• Based on a review of 151 formal and informal sources, providing a broad literature foundation.
• Includes insights from nine semi-structured interviews, adding qualitative depth.
• Categorizes pipeline tasks into four distinct stages: Data, Model, Dev, and Ops.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

Monika Steidl, Universität Innsbruck

Collection Method

Multivocal literature review and nine semi-structured interviews