The Analytics Power Hour #286: Metrics Layers. Data Dictionaries. Maybe It's All Semantic (Layers)? With Cindi Howson
12 snips
Dec 9, 2025 Cindi Howson, Chief Data & AI Strategy Officer at ThoughtSpot and a seasoned expert in business intelligence, discusses the resurgence of semantic layers in data strategy. She explains how these layers provide essential business context to raw data, reducing misunderstandings across departments. Cindi debates the need for multiple, context-driven semantic implementations rather than a single monolithic layer. She also highlights the importance of data modeling skills for analysts and offers practical tips for creating effective semantic layers.
AI Snips
Chapters
Transcript
Episode notes
Semantic Layer Purpose And Context
- A semantic layer maps business terms to physical data structures so non-technical users can query data without knowing SQL.
- Context matters: the same metric (e.g., revenue) can mean different things across functions like finance or supply chain.
Metrics Depend On Business Context
- Business context changes metric meaning, so definitions must capture origin and intended use.
- Different stakeholders (finance, sales) will interpret the same field differently and expect tailored definitions.
Use Semantic Layers To Improve AI Accuracy
- Build semantic layers to provide context for LLMs and reduce hallucinations in AI-driven queries.
- Use clear business definitions so AI and users generate accurate, trustworthy answers.