Each night the city enriches parcel records with mobility indices before morning planning views refresh. Where should that enrichment job live in a layered Big Data architecture?

Nightly enrichment of parcel records with mobility indices is a scheduled batch transformation that belongs in the processing layer of a layered Big Data architecture. Processing prepares analytical products; interactive serving then exposes morning planning views without embedding heavy joins inside UI click paths. Keeping enrichment in processing preserves repeatability, observability, and clean separation from clerk-facing consoles. Running enrichment only inside the dispatcher console’s real-time UI click handlers would stall operations and couple batch logic to interactive latency budgets. Performing it solely on citizens’ phones as client-side spreadsheet formulas cannot produce authoritative citywide planning inputs or auditable lineage. Manual copy-paste into a council agenda Word template is neither scalable nor acceptable as architecture for recurring municipal enrichment. Exam caveat: batch processing is not serving; candidates often misplace enrichment in the UI or assume the lake alone transforms data without a processing stage. Operational check: schedule the parcel–mobility job in the processing orchestration tool, land outputs in curated tables or serving marts before morning refresh, alert on failure before briefings, and keep dispatch UI read-only against enriched products rather than recomputing them per click.