What This Is
Semantic Stack is a resource for designing better data systems in the age of AI.
It combines:
- Best practices → practical, downloadable guides (Markdown) for data modeling and semantic layers
- Strategy → clear thinking on how data architectures evolve with AI
- Tools → a lightweight explorer to understand data before modeling
The Problem
AI is making data work faster—but not better.
Most teams still struggle with:
- unclear data structures
- unreliable joins and metrics
- fragmented business logic
- low trust in outputs
AI doesn’t fix this. It amplifies it.
The Approach
Semantic Stack is built on a simple principle:
You can’t model, govern, or use data with AI until you understand it.
Three focus areas:
- Data Understanding → profiling, relationships, structure
- Semantic Modeling → consistent metrics and business logic
- AI Context → adding meaning, documentation, and relationships for reasoning
What You’ll Find Here
- Downloadable best practices (MD)
Clear, reusable frameworks for:
- data modeling
- semantic layers
- data quality and joins
- Strategy & perspective
How to design data stacks that work with AI:
- semantic layer as control plane
- measurement vs meaning (metrics vs relationships)
- reducing complexity and increasing trust
- Tools
A local, in-browser explorer to:
- profile datasets
- detect joins and keys
- generate context for modeling
Point of View
- Semantic layers are more important with AI, not less
- Data modeling is shifting from SQL writing → data understanding
- The biggest bottleneck is trust, not computation
Who This Is For
Data professionals who want to build:
- reliable metrics
- clear data models
- AI-ready data systems