Section 11

System Design Flow

The canonical pipeline shape used in every deployment in this playbook.

Every system has the same five stages. Only the technology underneath changes.

Source
API / DB / files
Ingestion
raw landing
Bronze
raw_*
Silver
_clean
Gold
_summary
Outputs
BI / API / files

The layers

Source

External systems you don't control: APIs, OLTP databases, SFTP drops, flat files, webhooks. Treat as read-only.

Ingestion

A thin job that pulls from source and lands raw bytes/rows in the bronze layer. Zero transformation. Stamp every row with _loaded_at and _source.

Bronze — raw

Append-only mirror of source. Same column names, no business logic.

Silver — cleaned & conformed

Type-cast, deduplicated, joined to reference data, business keys resolved. One row per logical entity.

Gold — business-ready

Aggregates, KPIs, dimensional models. The only layer your dashboards and APIs read from.

Outputs

BI tools, REST APIs, Excel exports, reverse-ETL into operational systems.

Rule of thumb
If a stakeholder asks "where does this number come from?", you should be able to trace gold → silver → bronze → source in under a minute.