AI SIGNALS™ generates, organises, validates, and maintains the 167 atomic machine-readable signals that AI systems use to discover, classify, verify, and recommend businesses. The structured signal layer is the difference between being on a website and being in the AI knowledge graph.
Most businesses understand that content matters for search engines. Far fewer understand that AI systems do not primarily read content — they read signals. Structured, atomic, machine-readable data points that tell AI systems precisely what a business is, what it does, who it serves, what it has proven, and whether it can be trusted.
These signals are not embedded in website copy. They are deployed in structured files, schema markup, edge-delivered JSON, and linked data graphs. A business without a signal layer is not just less visible — it is categorically different in how AI systems process and weight it.
AI SIGNALS™ builds and maintains the complete signal layer — 167 signals across three categories — ensuring every signal is present, correctly structured, consistently maintained, and accessible to every AI system that queries it.
Each category targets a distinct AI system behaviour. All three must be complete for full signal coverage.
Signals that establish who the business is at the entity level. Legal name, all name variants, category, NACE/SIC code, founding date, headquarters, all locations, canonical URL, social profiles, registration number, VAT, entity type, entity status, parent entity, subsidiaries, and all sameAs references to authoritative external profiles.
Signals that prove the business is who it claims to be. All certification references with issuing body URLs, awards, professional body memberships, case study schema, client testimonials, third-party reviews, citation count, sector directory presence, SHA-256 proof files, OpenTimestamps anchors, and all verification statuses.
Signals that map the business to buyer intent and declare its governance posture. Primary and secondary service categories, all service-to-intent mappings, buyer personas, problem solved, delivery formats, pricing model, geographic and sector intent filters, negative intents, AI crawler permissions, and all governance file statuses.
Signal values are determined first — from the Business Knowledge Graph™. Files are generated from those signal values. Reversing this order produces inconsistent, error-prone deployments that create conflicting data in the AI knowledge graph. This rule is never violated.
Every signal passes through four validation stages. No unvalidated signal is ever deployed.
Is every required property present? Missing required fields are flagged before any file is generated.
Does this signal conflict with any other signal in the set? Cross-signal conflicts are resolved before deployment.
Can this signal's value be confirmed from an external source? Unverifiable signals are flagged for human review.
Is this signal accessible to the AI crawlers that need it? Deployment path verified before the signal goes live.
All 167 signals in a versioned registry with current status, last validation date, next review date, and source reference for every signal.
ai.json · entities.json · intents.json · ai-actions.json · ai-routes.json · ai-validation.json · allow-lane-matrix.json · adn.json — all generated from the signal set.
Structured report of every signal's validation status — completeness, consistency, verifiability, and accessibility — with findings and corrective actions for every failing signal.
Real-time view of signal health across all 167 signals with trend data and anomaly alerts. Powered by AI SONAR™ once the signal layer is deployed.