Hermes Wiki

DataEng

Data engineering concerns — pipelines, transformation, orchestration (see SYSTEM_DESIGN_ALIGNMENT.md section 9.14 for how this differs from backend engineering).

Why we need this / what value this brings

Once data needs to move through multiple stages (raw → transformed → served), pipeline thinking replaces simple query thinking.

When to use this

Once there's a real pipeline (e.g. feeding Analytics/DataWarehouse-BI) — not needed while all data lives in one operational DB.

How to use or implement this

Start with a simple, scheduled batch job before reaching for an orchestrator (Airflow) or streaming pipeline.

Research questions

  • Does Localz have any pipeline yet, or is this purely forward-looking for when Analytics/DataWarehouse-BI becomes real?

Empty folder — drop notes, links, and findings here as you research.

Hermes Wiki