Technical topic
Evolving semantic layers for data agents
Govern metric definitions, provenance, retrieval, semantic changes, and drift in data-agent workflows.
Direct answer
A data agent can execute a valid query and still answer the wrong business question when a metric or entity has the wrong meaning. This book makes definitions, provenance, retrieval, proposed updates, and release tests inspectable. It follows a bounded evolution loop: gather evidence from work, propose a semantic change, evaluate its impact on real questions, and admit only the revision that earns its release.
What this topic helps you decide
define business metrics before a data agent queries
Identify the business question and metric contract before authoring a query.
retrieve semantic definitions during agent work
Distinguish observed task failure from a justified semantic update.
detect drift in an evolving semantic layer
Release, monitor, and reverse definition changes without erasing prior meaning.
Practical questions answered
- How do I define a metric before a data agent queries it?
Specify a business metric’s population, event time, exclusions, aggregation, and owner before an agent translates a question into SQL.
- What provenance should a semantic-layer claim retain?
Attach source, transformation, owner, validity, and confidence to definitions so agents can inspect how a business claim was formed.
- How do I build an ontology from workflow evidence?
Derive a small ontology from decisions, records, exceptions, and competency questions rather than from a large noun inventory.
- How should a data agent retrieve definitions while working?
Resolve task-relevant semantic definitions by identity, time, scope, and provenance before an agent builds a data query.
- When should an agent trajectory propose a semantic change?
Distinguish a candidate gap in semantic definitions from retrieval, query, data-quality, and user-intent failures in agent traces.
- How do I evaluate a semantic-layer change before release?
Compare a proposed definition against source support, old and new business questions, data-agent outcomes, and rollback obligations.
- How do I detect drift in an evolving semantic layer?
Monitor source, definition, mapping, and usage changes separately so a data agent does not keep producing plausible answers from stale meaning.
Go deeper with a field guide
EvoOntology in Practice
Building and Evaluating Self-Evolving Semantic Layers for Data Agents
Explore EvoOntology in PracticeReusable resources
The practical guides include original decision tables, schemas, or diagnostic checklists where a reusable artifact improves the answer.