Diagrama de network sobre normalización de bases de datos

Tech architecture infographic titled "Database Normalization" using archetype DATA PIPELINE (extract → transform → load). Show a clear executive-friendly flow of how raw business data becomes normalized relational tables. Render labeled boxes connected by directional arrows in a simple left-to-right pipeline with supporting system components: Browser — "Executive Dashboard" — role: "Views source records and normalized outputs"; API — "Data Intake API" — role: "Receives structured business data via REST"; Queue — "Ingestion Queue" — role: "Buffers incoming records for processing"; Transform Engine — "Normalization Processor" — role: "Splits entities into 1NF, 2NF, 3NF tables"; DB — "Staging Database" — role: "Stores raw imported rows before cleanup"; DB — "Normalized Relational Database" — role: "Stores clean tables with primary and foreign keys"; Cache — "Query Cache" — role: "Speeds up repeated read requests"; Browser — "Reporting View" — role: "Displays consistent analytics from normalized data". Include optional small support boxes: "Data Quality Rules" — role: "Validates duplicates, nulls, and key integrity" and generic cloud icon box "External Source Systems" — role: "Provide CSV, ERP, CRM, or app data". Arrows must be labeled in English with realistic payloads: "CSV / JSON records", "HTTPS REST request", "enqueue job", "raw rows", "deduplicate + split entities", "create PK/FK relations", "SQL INSERT / UPDATE", "cached query result", "JSON response". Visually demonstrate normalization with compact example sub-boxes inside or near the processor: "Unnormalized Order Record", then split into "Customers", "Orders", "Order Items", "Products" tables, with labels like "remove repeating groups", "separate dependencies", "link by foreign key". Add a numbered legend 1-7 in English walking through lifecycle: 1. Source data is collected from business systems. 2. API accepts records and places them on the queue. 3. Raw data lands in staging for review. 4. Processor cleans fields and removes duplicate values. 5. Data is reorganized into normalized tables for 1NF, 2NF, and 3NF. 6. Relational database stores linked entities with primary and foreign keys. 7. Reports read consistent data through cache and API-backed views. Keep wording accessible for non-technical execs while technically accurate. Add small callout notes in English: "Goal: reduce redundancy", "Benefit: improve consistency", "Trade-off: more joins for reporting", "Example normal forms: 1NF, 2NF, 3NF". Visual style: hand-drawn whiteboard sketch, cool blue and cyan palette, soft marker lines, neat doodle icons, lightly shaded boxes, approachable boardroom-explainer mood, editorial developer-blog illustration, isometric or flat tech-diagram style, vector-clean infographic layout. No real cloud-vendor logos; use only generic cloud, database cylinder, browser window, gear, queue, and cache icons. All text MUST be written in English (array). Every heading, label, caption, legend and metric name in the image must be in English — not English. Spell each English word correctly using English characters and diacritics. Numbers stay as digits, no real cloud-vendor logos (AWS / GCP / Azure) — use generic cloud icons, no watermarks No real cloud-vendor logos (AWS, GCP, Azure) beyond generic cloud icons. Common protocol names (HTTPS, TCP, JWT, OAuth, REST, GraphQL) stay in canonical English form. No security-claim overstatements (do not present diagrams as audited reference architectures).