Knowledge Graph Store

Data storeKnowledge & DataKnowledge & DataVariation point (abstract)arc:KnowledgeGraphStore

A store of entities and relationships enabling multi-hop relational and causal reasoning that vector similarity cannot represent.

Responsibility. Stores entities and relations.

Also known as: Knowledge Graph, Graph Database, Graph database, Wikidata structured evidence, Episode relationship graph, Unified semantic layer, Shared semantic knowledge graph across paradigms

Direct neighbourhood (hover for relationship types)

Variants

Variant	When to choose
Property Graph Store	Choose when rapid development and schema flexibility (properties on nodes and edges, multiple labels, schema evolution) matter more than formal semantics; the industry-standard choice for agent systems.
RDF Triple Store	Choose when interoperability across systems requires standardized ontologies and formal W3C semantics (e.g., academic knowledge bases like Wikidata).

Relationships

is written by dependency

emits telemetry to dynamic

Metrics Collector Ch1.7B

is constrained by control

is guarded by control

Graph Consistency Validator Ch5.8

is audited by assurance

is monitored by assurance

SLO Monitor Ch1.7B

Design guidance

SHOULD be accessed through a pooled driver connection, since agents may issue dozens of graph queries per conversation turn.
SHOULD NOT serve single-node property lookups; key-value or document stores answer those faster.
MUST bound variable-length traversals with explicit relationship types and maximum depth, and index frequently queried properties.
SHOULD scale with clustering, sharding by entity type, and read replicas for query distribution.
SHOULD evolve schema additively and version relationship types when semantics change instead of deleting labels or properties.
SHOULD be scaled independently of agent compute, vertically first, then with read replicas, sharding or federation.
SHOULD be accessed through pooled connections and batched queries when many replicas query concurrently.
SHOULD optimize common traversal patterns via materialized paths or cached traversals; complex multi-hop traversals over millions of nodes become expensive.
SHOULD be chosen when reasoning requires explicit relationships, multi-hop inference or guaranteed consistency, accepting extraction and maintenance overhead.
SHOULD represent uncertain knowledge with confidence weights rather than binary triples in domains such as clinical knowledge.
SHOULD record which source system contributed each fact to provide lineage.

Quantitative guidance

As stated by the sources; verify before use.

A three-hop filtered Neo4j query ~80ms vs ~5ms for a key-value property lookup (Ch1.7B).
An unbounded variable-length path query may finish in seconds at 1,000 nodes but time out at 100,000 nodes (Ch1.7B).
Wikidata holds 100M+ entities (Ch5.8).
Multi-hop queries taking milliseconds on million-node graphs can take seconds or minutes on billion-node graphs without optimization (Ch5.8).

Classification

Patterns: Multi-hop traversalGraph + vector hybrid memoryMulti-hop reasoning via graph traversalPattern matching queriesRead replicasSharding / federated graphsConnection poolingQuery batchingHappened-before timelinesMulti-agent interaction topologyMaterialized paths / cached traversalsEntities, relations, properties and rulesTransitive multi-hop inferenceSharding / partitioning across serversConfidence-weighted (probabilistic) factsData lineage to source systemsProbabilistic edgesGraph evolution from learned discoveries
Technologies: Neo4jTigerGraphAmazon NeptuneCypherBolt protocolNeo4j clusteringRead replicasNeo4j EnterpriseWikidataDBpedia
Quality attributes: Explainability (NIST AI RMF: explainable and interpretable)Functional suitability: correctness and validity (ISO/IEC 25010 | NIST AI RMF: valid)Performance efficiency (ISO/IEC 25010)Reliability (ISO/IEC 25010 | NIST AI RMF: valid and reliable)
Risks mitigated: Broken multi-hop reasoning chains in vector-only RAGOpaque similarity-based answersContradictory facts

Sources

Ch1.4: T. Nguyen, "Memory and Perception Systems," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 1.4. ISBN: 9798244538229.
Ch1.7A: T. Nguyen, "Relational Reasoning with Knowledge Graphs - The Fundamentals, Integration, and Extraction," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 1.7A. ISBN: 9798244538229.
Ch1.7B: T. Nguyen, "Relational Reasoning with Knowledge Graphs - Hybrid RAG+KG Integration," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 1.7B. ISBN: 9798244538229.
Ch1.8: T. Nguyen, "Scalability and Production Deployment," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 1.8. ISBN: 9798244538229.
Ch3.3: T. Nguyen, "Web Navigation and Interaction Benchmarks," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 3.3. ISBN: 9798244538229.
Ch3.6: T. Nguyen, "Trace Analysis and Execution Debugging," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 3.6. ISBN: 9798244538229.
Ch5.7: T. Nguyen, "Episodic Memory," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.7. ISBN: 9798244538229.
Ch5.8: T. Nguyen, "Semantic Memory," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.8. ISBN: 9798244538229.
Ch5.13: T. Nguyen, "Hybrid Decision Systems Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.13. ISBN: 9798244538229.
Ch6.3A: T. Nguyen, "ETL Pipeline Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 6.3A. ISBN: 9798244538229.
Ch6.4: T. Nguyen, "Data Quality Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 6.4. ISBN: 9798244538229.