Define your KPIs, resolve conflicts, monitor drift, and give every team (and every AI) a single governed source of truth.
Most enterprises try to govern data meaning through a mix of documentation, wiki pages, dbt descriptions, and word of mouth. It works, until a business event changes a definition and there's no system that knows.
The Semantic Manager is the workspace where your semantic layer is built, maintained, monitored, and kept accurate over time. Not a one-time export. A living governance environment.
Six capabilities, built to work together, from initial definition capture through continuous drift detection and team collaboration.
Create, cluster, and version your critical semantic units. KPIs, metrics, business rules, domain terms. Every definition carries its logic, applicable scope, and full change history. Business events update definitions, not orphan them.
Versioned · ScopedEvery definition is checked against your existing knowledge base before it's saved. Overlapping definitions, contradictory logic, ambiguous terms, flagged immediately, with plain-language explanations. This is how you stop hallucinations at the source.
Pre-save · AutomaticFinance, marketing, and sales don't always agree on what "revenue" means. Define domain-specific and role-specific interpretations of the same underlying metric, without forking models. Every AI query knows which interpretation applies.
Multi-perspective · No forksA visual overview of your semantic layer, where definitions are solid, where conflicts exist, where drift is occurring. Think uptime monitoring, but for the meaning layer of your AI stack. Spot blind spots before they reach production.
Real-time · VisualSemantic drift happens when the world changes and your knowledge layer doesn't. LazyFox monitors usage signals, how queries are phrased, where users push back, what gets corrected, and surfaces drift as it emerges, not after damage is done.
Continuous · Signal-drivenAnyone can propose a definition or flag a conflict. Approvals, edit rights, and visibility are configurable at the unit, group, or role level. No emailing engineers to merge a PR that a business analyst wrote. Governance the whole organization can participate in.
Team-wide · ConfigurableWhen two people define the same KPI differently, AI doesn't pick the right one, it picks whichever has the higher probability in the vector database. That's where hallucinations come from. Not missing data. Conflicting definitions stored in the same space.
Every time a semantic unit is created or edited, the Semantic Manager validates it against your full knowledge base before it's saved. Conflicts must be resolved. Ambiguities can be acknowledged or refined. Nothing lands without passing a gate.
The health map gives you a visual overview of your semantic layer. Not just what you've defined, but how it's holding up. Where usage is high. Where drift is emerging. Where conflicts have gone unresolved.
Most organizations have no visibility into the health of their semantic layer. They find out something is wrong when an AI query returns a wrong answer. LazyFox shows you the signal before that happens.
The Semantic Manager is built for the whole organization, not just the data team. Every role gets what they need without stepping on each other's work.
Every definition you manage in the Semantic Manager flows into all connected systems, consistently, without sync jobs or manual export.
See all integrations →We'll walk you through the Semantic Manager and show you how your team's highest-value KPIs would look as governed semantic units.