apd-data-modeling-decision
| apd-data-modeling-decision | |
|---|---|
| Version | 1.0 |
| Category | Analytics Engineering |
| Type | Decision assistant |
| Cost / run | TBD |
| Best models | TBD |
Analytics Engineering › apd-data-modeling-decision
Guides selection of a data modeling approach for a specific client environment (Data Vault, Inmon/3NF, Kimball, star/snowflake, One Big Table, lakehouse raw-plus-marts) and produces a decision memo: options considered, recommendation, tradeoffs, and impacted assets. Built from issue #28.
Available as $apd-data-modeling-decision / /apd-data-modeling-decision (source · tests).
How to use
- Invoke the skill with the client environment context: platform, team skills, data volumes, existing assets, reporting needs
- Answer its clarifying questions; the recommendation quality tracks the input quality
- Take the decision memo into the stakeholder conversation; it's built to be walked through
Help me choose a modeling approach for a client on Snowflake + dbt.
~40 sources, small AE team (2 people), heavy self-serve BI demand,
no existing warehouse conventions. Produce the decision memo.
Tips
| Do | Why |
|---|---|
| Give real constraints, not ideals | Team size and skill mix often decide between Vault and Kimball, not theory |
| Use the memo as the deliverable | It's structured for stakeholder walkthroughs; don't rewrite it from scratch |
| Revisit when the environment changes | A platform migration or team change can flip the recommendation |
Related skills
| Skill | Relationship |
|---|---|
apd-ae-content-reviewer |
If the decision memo becomes public thought leadership, run it through review |
| Kimball / Data Vault best-practice skills | Proposed (#29, #30); downstream of the decision this skill makes |