Data Warehouse Projects: A Short Course for IT Executives

Explore data warehousing terminology, including data model, dimensional modeling, star and snowflake schemas, dimension and fact tables, slowly changing dimensions, and data contracts.

Discover why organizations need a data warehouse to separate transactional and analytical workloads. Prevent data silos, enable consistent reporting, and support 360 analysis with master data management.

Build a self-service analytics platform by implementing a data warehouse that cleanses, moves, and conforms data through ETL into centralized historical and reporting databases for effective analytics.

Clarify data terminology by distinguishing data warehouses, data marts, and data lakes. Explain when a data mart or data lake fits a project while highlighting data retrieval and data structures.

Choose a database by use case rather than comfort with existing tech, then design a dedicated data warehouse box and implement appropriate orchestration and reporting tools.

Explore two design approaches for data warehouses, emphasizing dimensional modeling over normalization, using a Kimball-inspired, question-driven method to build cubes from a central data source.

Visualize a data model to document and socialize a data warehouse project, showing tables, relationships, and star schema elements like fact and dimension tables.

Understand dimension tables in a dimensional model, including date dimension and degenerate dimensions, slowly changing, conformed and role play dimensions, outriggers, and how they connect to fact tables.

Explore fact tables in data warehouses, distinguishing additive, semi-additive, and non-additive facts; learn how cumulative and snapshot facts, primary keys, and dimension combinations ensure unique, accurate records.

Learn how master data management ensures a single, clean set of key entities by identifying duplicates, aligning local and home office names, and coordinating data stewards and processes.

Explore analytics maturity models as tools to measure business analytics capability, noting no industry standard, from reporting at the base to advanced levels, and learn to move progress upward.

Explore ETL processes for moving data into a data warehouse, including cleansing, conforming, and building disciplined pipelines with staging tables, a common model, and data contracts.

Explain data contracts as formal agreements governing data transfers into the data warehouse, emphasizing change control, interdepartmental coordination, and how to design resilient pipelines with buffer databases and defensive programming.

Develop a standardized ETL framework to move data into and out of the warehouse, handling internal and external sources and multiple file types for consistent diagnostics and cost savings.

Discover the mass street ETF framework, an open source MIT license batch processing tool for relational databases, with Python and sequel guidance, data contracts, and people as a focus.

Drive organizational buy-in for a data warehouse by rallying all departments, running a road show from the CEO, and delivering quick wins.

Establish a data governance function from day one to protect the warehouse as a trusted system of record. Create a central authority with a cross-functional team to enforce data cleanliness.

Identify the data steward as the focal point of the human-in-the-loop MDMA process, balancing overnight batch match conflicts with practical, department-aligned governance.

Document and standardize data flows and schemas to empower self-serve analytics. Create and maintain a data dictionary with descriptive column names and business logic explanations, ensuring read-only access.

Wraps up a concise, comprehensive data warehouse course with guidance from a data engineer, notes on organizational gaps, and an invitation to future data science, engineering, and analytics courses.