The current pipeline only lints Python and then deploys whatever tar.gz is sitting in a shared bucket. A new commit never proves the model is still fit to ship. What CI/CD principle is missing for ML?
Select an answer to reveal the explanation.
Short Explanation
The pipeline only lints Python and ships whatever tar.gz sits in a shared bucket. Every change must be built, tested (including a model-artifact or evaluation gate), and only then delivered. App lint is not an ML pipeline.
Full Explanation
CI/CD applied to ML means every change is built, tested (including the model artifact or an evaluation gate), and only then delivered. App lint plus a shared-bucket tarball is not that pipeline. Personalize is not the gate.