Designing and Implementing Semantic Data Layers

Gartner predicts that by 2030, semantic layers will be viewed as critical infrastructure, right alongside your data platform and cybersecurity. That is an attention-grabbing claim, and most data professionals are not yet sure what it means for them. Here is what it means. The data problems that you face every day are problems of meaning: data silos, data disparity, integration complexities, data distrust, confident but wrong AI answers. These may look like technical problems, but the root causes are found in inconsistency of names, definitions, interpretations, and assumptions across systems, databases, and use cases. Your infrastructure works. What’s missing is shared, explicit, governed meaning that no platform or technology will provide.
Many organizations and practitioners look at semantic layers and conclude that they are too specialized, too academic, or simply out of reach. That reaction comes from treating them as something exotic. Dave Wells’ new book, Designing and Implementing Semantic Data Layers, is written to overcome those barriers with practical and accessible guidance. Ultimately, semantic data layers are software. Building them should follow proven software development practices: start with business needs, then architectural placement, followed by design, coding, testing, and deployment. In this session, we look at the full progression from why meaning matters to where semantic layers are needed, and then to how the work actually gets done.
Olga Maydanchik captures the essence of these ideas in her Amazon review:

"I love how the book moves from conceptual discussions into the 'how'...
All of a sudden, you can see how to break the work down into manageable pieces."
You will see:

• Why familiar data problems are semantic problems that technology alone can’t solve
• Why “the semantic layer” is flawed thinking and why you need a purposefully designed network of multiple semantic layers
• Five types of semantic data layers – enterprise, domain, integration, enrichment, and consumption – and the role of each
• A visual, business-friendly modeling technique that makes ontology and taxonomy approachable without starting in a formal modeling language
• Semantic layer implementation patterns and how to choose among them
• How to get started, and why you should follow the value instead of the org chart