Appendix - Cloud Platforms

What We Learned

Map model access, retrieval, guardrails, agents and data services across cloud platforms.
Understand Databricks' lakehouse, governance, experiment tracking and distributed data processing.
Choose managed services or self-managed infrastructure from workload constraints and evidence.
Compare identity, data location, quotas, cost, portability and operational ownership.
Treat vendor names as a dated snapshot; compare capabilities before selecting a platform.

Term	Short meaning
AWS / GCP	Amazon Web Services / Google Cloud Platform: cloud infrastructure and managed services.
Managed service	Provider operates a capability; the customer still owns configuration and application behavior.
Model catalog / endpoint	Available model inventory / deployed model access interface.
RAG	Retrieval-Augmented Generation: answer using retrieved source material.
Guardrail / agent runtime	Content/action checks / infrastructure that executes agent runs.
Lakehouse	Data architecture combining lake storage with warehouse-style management.
Medallion architecture	Bronze raw, silver cleaned and gold consumption-ready data layers.
Delta Lake	Table format and tooling for reliable, transactional data-lake tables.
Unity Catalog	Databricks governance for data and AI assets.
MLflow / MLOps	Experiment and model lifecycle tooling / Machine Learning Operations.
Spark / PySpark	Distributed data processing engine / its Python interface.
DBU	Databricks Unit: a platform usage unit used in pricing.
GKE / TPU	Google Kubernetes Engine / Tensor Processing Unit: managed Kubernetes / Google's ML accelerator.
Vendor lock-in	Switching cost caused by provider-specific interfaces, data or infrastructure.