Question

A data science team runs dozens of model prototyping iterations in Vertex AI Workbench notebooks. Each run produces a training dataset snapshot, a preprocessing script, hyperparameters, and a model artifact stored in Cloud Storage. Months later, an auditor asks the team to show exactly which dataset version and parameters produced a specific model that was later promoted to production. The team currently only logs metrics to a spreadsheet. What is the most effective way to reliably capture this lineage going forward with minimal custom tooling?

Accepted Answer

Use Vertex ML Metadata via Vertex AI Experiments to log each run's parameters, metrics, and register input/output artifacts so lineage between datasets, executions, and models is automatically tracked and queryable.

Answer

Store all datasets, scripts, and models in a single BigQuery table with a timestamp column and query it later to reconstruct which artifacts were used.

Answer

Enable Cloud Storage object versioning on the model bucket and rely on the version history to determine which dataset produced each model.

Answer

Save each notebook as a git commit and use the commit message to record the dataset version and hyperparameters for every run.