# Smaddanki

> Writing on agent data layers, silent failure, risk and controls, the real TCO of agents, and the changing data function.



## Categories

- [Data pipelines](https://smaddanki.com/categories/data-pipelines/): Design principles, components and building blocks for the pipelines everything else sits on.

## Articles

- [Core data pipeline building blocks every engineer should know](https://smaddanki.com/writing/core-data-pipeline-building-blocks-every-engineer-should-know/): Seven decisions — processing, ingestion, loading, change management, processing method, quality and storage — determine what a pipeline can do.
- [Data pipeline components: architectural blueprint and choices](https://smaddanki.com/writing/data-pipeline-components-architectural-blueprint/): Every data pipeline is the same set of components; what differs is which implementation you pick for each.
- [13 essential data pipeline design principles for effective data engineering](https://smaddanki.com/writing/13-design-principles-for-data-pipelines/): A pipeline is reliable because of thirteen properties it holds, not because of the tools it is built from.