A team wants retraining to start automatically whenever a new data batch passes validation, with a new model promoted to an endpoint only if evaluation thresholds are met. Which pairing fits this continuous training and delivery pattern?
Show answer and explanation
Answer: B
Cloud Build can start automated workflows on code or data events, and Vertex AI Pipelines runs the training, evaluation, and deployment steps as tracked stages. Conditional gates in the pipeline let promotion happen only when metrics pass. Together they automate continuous training and delivery with the required controls.
Why the other options are wrong
- A. Manual DNS edits are unrelated to model training and cannot evaluate or promote anything. They do not automate any part of the ML lifecycle.
- C. Change streams capture database modifications but do not orchestrate training or deployment. A spreadsheet has no role in gating model promotion.
- D. Manual uploads bypass automated validation and evaluation gates. The scenario calls for repeatable, automated retraining and promotion, which manual steps cannot provide.