This five-course specialization takes you from lakehouse fundamentals to production-grade AI systems on the Databricks platform. You begin by building data pipelines with Apache Spark and Delta Lake, learning medallion architecture (bronze, silver, gold) and Unity Catalog governance. You then advance to Delta Live Tables for declarative ETL with built-in data quality expectations, streaming ingestion with Auto Loader, and Change Data Capture with APPLY CHANGES. The specialization progresses into machine learning engineering with MLflow tracking and the Databricks Model Registry, generative AI with LLM fine-tuning and RAG pipelines using Vector Search, and concludes with production governance — model serving, A/B testing, monitoring, and CI/CD for ML systems. Every course includes hands-on labs on the Databricks platform using real-world datasets and production patterns.

Übungsprojekt

Across the specialization, you build progressively complex systems on Databricks. In Course 1, you construct an end-to-end medallion pipeline (bronze to silver to gold) with Delta Lake MERGE operations and Databricks Workflows orchestration. In Course 2, you build a production Delta Live Tables pipeline with expectations-based data quality, streaming ingestion via Auto Loader, and Change Data Capture for an inventory management system. Later courses extend this foundation with MLflow experiment tracking, model registration, LLM fine-tuning, retrieval-augmented generation, and automated model serving with governance controls. Each project uses the Databricks Community Edition or workspace — no cloud billing required for the labs.