DRI: @eduardobonet
MLOps is a Single-Engineer Group within our Incubation Engineering Department. This group works on early feature exploration and validation related to the MLOps group within the ModelOps stage.
Mission: Make GitLab a tool Data Scientists and Machine Learning Engineers love to use.
Vision: Identify opportunities in our portfolio to explore ways where GitLab can provide a better user experience for Data Science and Machine Learning across the entire Machine Learning life cycle (model creation, testing, deployment, monitoring, and iteration).
May 1st 2023: Model 15.11 Overview
Model Registry & Model Experiments
Mission: Make it dead simple for Data Scientists to manage their model lifecycle within GitLab, from testing different candidates to, packaging a new model version.
More on ML Experiment Tracking
MLops is a large and young field. We used Jobs to be Done to make it explicit what are the problems users expect MLOps to solve:
We are keeping a backlog for potential exploration areas. Anyone is welcomed to pitch in new ideas using the Backlog Epic
| Reach | Impact | Confidence | Effort | Colab | MLOps Branding* | RICE+ | |
|---|---|---|---|---|---|---|---|
| JupyterLab-GitLab Plugin | 2 | 2 | 2 | 2 | 1 | 3 | 24 |
| Enable GitLab Runners for ML Use cases | 3 | 3 | 3 | 2 | 2 | 1.5 | 13.5 |
| Analysis Repository, GitLab Pages for Data Science | 1 | 2 | 1 | 2 | 1 | 3 | 6 |
| Improve pipeline usage for ML Use cases | 3 | 1 | 3 | 3 | 3 | 1 | 2 |
| DVC Integration/Data Registry | - | - | - | - | - | - | - |
