06-reference

practical data modeling mma ch12 synthesis

Wed Apr 08 2026 20:00:00 GMT-0400 (Eastern Daylight Time) ·reference ·source: Practical Data Modeling (Substack) ·by Joe Reis
data-modelingsynthesismixed-model-artsfive-campsintegrationchapter-12

Stringing Together the Building Blocks (Ch 12)

Final chapter of Part 2. Short but pivotal — Reis shifts from isolated building blocks to showing how they form a system. Grain changes force aggregation adjustments; temporal complexity demands time-awareness; conflicting semantics undermine entity models.

Seven-Question Synthesis Checklist

A portable tool for any model spanning multiple data forms: (1) What is the business question? (2) What data forms are involved? (3) Where do entities converge? (4) What is the target grain? (5) How do time dimensions align? (6) What is the semantic bridge? (7) What is the implementation pattern? Each question maps to a prior chapter’s building block.

Worked Example: Customer Support

Walks through a complete model across all five camps simultaneously. Relational (normalized entities with foreign keys), Analytics (wide fact table with dimensions), Application (nested documents), ML/AI (vector embeddings for ticket clustering), Knowledge (semantic network with typed edges). One business problem, five complementary approaches.

The example hits every building block: many-to-many ticket-product relationships need junction entities, bitemporal modeling handles priority escalations for SLA tracking, aggregation stress-tests enforce decomposable metrics (store numerator/denominator separately for ratios), and semantic clarity requires agreed definitions of “resolved.”

RDCO Relevance

The synthesis checklist is a practical framework for dbt project kickoffs — especially questions 4-6 on grain, temporal alignment, and semantic bridging. Cross-ref SDG pipeline architecture articles.