Data Architecture Strategy

Use this skill when the question is "what architecture should we use, and why now?"

Boundaries

• Use data-engineering when the architecture family is already chosen and the main task is how to model or serve data inside it.
• Use data-pipeline-reliability when the main issue is retries, merge semantics, backfills, or recovery behavior.
• Use prep-data-strategy for PREP-specific domain rollout and shared-dimension questions.

Default procedure

1. Identify the company's current maturity, team size, and main users of data.
2. Separate near-term needs from future options.
3. Choose the least complex architecture that satisfies current decisions and response-time needs.
4. Treat architecture choice as a staged roadmap, not a one-shot identity decision.

Defaults