Design a Robust ETL Pipeline Architecture for Your Data Platform

You are a senior data engineer with 10+ years of experience designing scalable ETL (Extract, Transform, Load) pipelines. I need you to design a comprehensive ETL pipeline architecture for the following scenario: **Data Context:** - Source systems: [LIST_OF_SOURCE_SYSTEMS, e.g., PostgreSQL, REST APIs, CSV files] - Target destination: [TARGET_DATA_WAREHOUSE, e.g., Snowflake, BigQuery, Redshift] - Data volume: [ESTIMATED_DAILY_DATA_VOLUME, e.g., 50GB/day] - Refresh frequency: [BATCH_OR_STREAMING, e.g., hourly batch, real-time streaming] - Primary use case: [BUSINESS_USE_CASE, e.g., customer analytics dashboard, financial reporting] **Please provide the following in your design:** 1. **Architecture Diagram Description**: Describe the end-to-end pipeline flow, including each component and how they connect. Specify which tools or services you recommend (e.g., Apache Airflow, dbt, Kafka, Fivetran) and justify each choice. 2. **Extraction Layer**: Detail the extraction strategy for each source system, including connection methods, incremental vs. full load logic, and change data capture (CDC) considerations. 3. **Transformation Layer**: Outline the transformation steps including data cleansing, deduplication, schema mapping, business logic application, and dimensional modeling approach (star schema, snowflake schema, or OBT). 4. **Loading Strategy**: Specify the loading pattern (upsert, append, truncate-and-reload), partitioning strategy, and indexing recommendations for the target warehouse. 5. **Error Handling & Monitoring**: Design a robust error-handling framework including retry logic, dead-letter queues, data validation checkpoints, alerting mechanisms, and logging standards. 6. **Scalability & Performance**: Address how the pipeline handles data volume growth, parallel processing, resource optimization, and backfill scenarios. 7. **Data Governance**: Include lineage tracking, metadata management, access controls, and PII handling procedures. Format the output with clear headers, bullet points, and include a sample DAG or workflow definition in pseudocode where applicable.