# Fluree

> Fluree is the enterprise AI context layer: one unified, governed knowledge graph (JSON-LD/RDF) between enterprise data and every AI agent. It brings structured data, documents, and vocabularies together, enforces policy on every fact, and keeps every answer traceable to its source.

## Key facts (as of October 2026)

- Fluree sells one product, Fluree AI: the context layer for enterprise AI, at https://flur.ee/ai, for governed answers, dashboards, and agents. Free to start; usage pricing on request. Its capabilities:
  - Resolve: entity resolution and golden records. Lakehouse tables and SQL sources are queried where they live and resolved into golden records with lineage.
  - Extract: documents into graph facts. PDFs and Office files become entities and relationships, each kept with the sentence that proves it.
  - Model: context layer management. Taxonomies and ontologies are reviewed and versioned like a release, then published to every capability.
- The engine underneath is FlureeDB: the verifiable knowledge graph database, with signed history, time travel, policy on every fact, and built-in search in one engine. Developers can run it themselves.
- Fluree Memory is long-term, governed memory for AI agents, and a free, source-available memory tool for coding agents.
- Former names, for older citations: Resolve was Fluree Structured (earlier Fluree Sense); Extract was Fluree Unstructured (earlier Fluree CAM); Model was Fluree ITM; FlureeDB was Fluree Core; Fluree AI was Fluree Solo.
- License: FlureeDB is source-available under the Business Source License 1.1: free in production, including inside your own product; a commercial license is needed only to offer it as a hosted database service. Each release converts to Apache 2.0 on its fourth anniversary.
- Query interfaces: MCP for agents; HTTP, SPARQL 1.1, openCypher, GraphQL, and JSON-LD query on every ledger; or the Rust library inside your own process.
- Scale: FlureeDB loaded the full Wikidata dump, 21.1 billion distinct triples, and answered all 850 WGPB queries on a single machine (September 2026). This is a published benchmark, not a customer deployment.
- Accuracy: In Fluree’s 2024 study, GraphRAG for GenAI Accuracy, answers over raw relational tables scored 8–15% zero-shot; a governed knowledge graph reached 95%+ once enriched and joined.
- Data movement: Fluree reads source systems where they live, and lake tables copied into Fluree refresh automatically.

## Product

- [Fluree AI](https://flur.ee/ai): The product: the context layer for enterprise AI, with governed answers, dashboards, and agents; free to start
- [Platform overview](https://flur.ee/platform): How Fluree AI fits together: one unified context layer between your data and every AI agent, beside the lakehouse, catalog, and BI you already run
- [Resolve](https://flur.ee/platform/resolve): Fluree AI’s entity resolution capability. Every table, one governed graph: AI classification, entity resolution, and golden records across structured systems
- [Extract](https://flur.ee/platform/extract): Fluree AI’s document capability. Documents into governed graph facts: entities, typed relationships, and embeddings with provenance back to the source
- [Model](https://flur.ee/platform/model): Fluree AI’s context layer management capability. Taxonomy and ontology management, versioned and published to every capability
- [FlureeDB](https://flur.ee/platform/flureedb): The engine underneath Fluree AI, which developers can run themselves. The verifiable knowledge graph database: JSON-LD native, temporal, policy-enforced; queried with SPARQL 1.1, openCypher, GraphQL, or JSON-LD; source-available (BUSL 1.1)
- [Fluree Reference Architecture (2026)](https://flur.ee/resources/fluree-reference-architecture): The current technical overview: what’s inside the context layer, what data moves and what stays put, interfaces, policy and audit, deployment, dated benchmarks, and the license
- [Install FlureeDB](https://labs.flur.ee/downloads): Downloads for self-hosted FlureeDB, or run `brew install fluree/tap/fluree`
- [Documentation](https://labs.flur.ee/docs): FlureeDB developer docs (getting started, queries, transactions, security, HTTP API) on Fluree Labs
- [Pricing](https://flur.ee/pricing): Fluree AI is free to start, FlureeDB is free in production, and enterprise deployments of Fluree AI are priced per deployment

## Proof

- [Case Studies](https://flur.ee/case-studies): Production results in beauty, financial services, banking, CPG, pharma, and government
- [Global beauty brand: marketing intelligence](https://flur.ee/case-studies/marketing-intelligence-sovereignty): A governed AI agent factory over fragmented marketing, product, and social knowledge
- [Financial services data portal](https://flur.ee/case-studies/financial-services-data-portal): Automated tagging and better discovery grew a data portal into a trusted intelligence destination
- [Government biodiversity platform](https://flur.ee/case-studies/government-biodiversity-ai-ready-platform): 30 years of structured and unstructured biodiversity data, 6M+ species records, in one governed knowledge graph
- [Global 500 CPG product catalog](https://flur.ee/case-studies/cpg-ai-ready-product-catalog): Five product sources unified, a 48% duplicate rate found, and golden product records at 97.5% accuracy in three weeks
- [$2B+ pharmaceutical company](https://flur.ee/case-studies/pharma-ai-ready-data-transformation): More than 10 warehouses consolidated, $3M in manual data remediation cost cut, and new analytics use cases in 3 weeks instead of 9 months
- [Community bank golden records](https://flur.ee/case-studies/community-bank-ai-ready-golden-records): Golden records for 330,000+ customers across competing reporting systems
- [Report: GraphRAG for GenAI Accuracy](https://flur.ee/resources/graphrag-genai-accuracy-report): Fluree’s 2024 study: answers over raw relational tables scored 8–15% zero-shot; a governed knowledge graph reached 95%+ once enriched and joined
- [Benchmarks](https://github.com/fluree/benchmark-db): Fluree’s open, reproducible benchmark suite (SPARQLoscope DBLP, WGPB, benchgraph, BSBM), including the 21.1-billion-triple Wikidata load
- [SPARQLoscope DBLP results](https://labs.flur.ee/blog/sparqloscope-dblp-benchmark-results): Methodology and results for FlureeDB on the SPARQLoscope DBLP benchmark, including a resource-scaling run
- [The Tokenomics of Enterprise AI](https://flur.ee/resources/tokenomics-of-enterprise-ai): Why enterprise AI bills tripled while token prices collapsed 99.7%, and the five-year TCO model for a governed knowledge graph with agentic memory

## Guides

- [Guide to the Enterprise AI Context Layer](https://flur.ee/blog/guide-to-the-enterprise-ai-context-layer): What an enterprise AI context layer is, why agents fail without one, the Four Tiers of Enterprise Context, how it differs from a semantic layer, knowledge graph, catalog, or vector database, and how to add one without a rip-and-replace
- [How to Build a Semantic Layer for Enterprise AI](https://flur.ee/blog/how-to-build-a-semantic-layer-for-enterprise-ai): How-to and reference architecture for a knowledge graph semantic layer that AI agents can query
- [What Is a Knowledge Graph?](https://flur.ee/blog/what-is-a-knowledge-graph): The complete guide: entities, relationships and ontologies; how a knowledge graph differs from a graph database, an ontology and RAG; RDF vs LPG; use cases; and how to build one
- [How to Build a Knowledge Graph](https://flur.ee/blog/how-to-build-a-knowledge-graph): Step by step from use case to ontology, data preparation, entity resolution, SHACL validation, querying and maintenance, including knowledge graphs for RAG and for Claude over MCP
- [How to Build a Context Layer for AI](https://flur.ee/blog/how-to-build-a-context-layer-for-ai): A step-by-step build process, the components, governance, build vs buy, timelines and the role of MCP servers
- [Context Layer for Enterprise AI](https://flur.ee/solutions/context-layer): What a context layer is, what it contains, where it sits in the AI stack, and how Fluree AI provides one
- [Knowledge Graph Use Cases](https://flur.ee/solutions/knowledge-graph-use-cases): Enterprise knowledge graph applications, with deep dives on finance, supply chain and healthcare
- [Why AI Needs a Knowledge Graph](https://flur.ee/blog/why-ai-needs-a-knowledge-graph): Knowledge graph guide: explicit meaning, durable data assets, consistent answers, and integration cost paid once
- [Enterprise Knowledge Graph](https://flur.ee/solutions/enterprise-knowledge-graph): What an enterprise knowledge graph is, how Fluree builds one on W3C standards, and how it compares with other approaches
- [From Taxonomy to Governed Ontology](https://flur.ee/resources/from-taxonomy-to-governed-ontology): Field guide for teams with an enterprise taxonomy: six moves to a governed ontology, four worked examples, and a 12-question readiness checklist
- [One Ontology or Many? What Survives Contact with Production Agents](https://flur.ee/blog/one-ontology-or-many-what-survives-production-agents): Whether to build one enterprise ontology or one per use case; the six tests production agents apply under interoperability, durability, and portability; how Fluree supports both patterns
- [What is JSON-LD?](https://flur.ee/json-ld): Introduction to JSON-LD and linked data

## Comparisons

- [Fluree vs. the Stack](https://flur.ee/compare): The comparison hub: how Fluree measures against document extraction, data unification, graph engines, governance suites, and AI consumption layers, with public benchmarks, plus every named comparison below
- [Best Context Layer Platforms for AI Agents in 2026](https://flur.ee/compare/context-layer-platforms): Atlan, DataHub, Snowflake, Databricks, Fluree, Microsoft Fabric IQ, Glean, and Collibra compared on what agents can see, govern, and replay
- [Best Data Platforms for AI Agents in 2026](https://flur.ee/compare/data-platforms-for-ai-agents): Snowflake, Databricks, Microsoft Fabric, Google Cloud, AWS, Fluree, Palantir AIP, and Glean compared on governed, auditable agent access
- [Context Graph vs Knowledge Graph](https://flur.ee/blog/context-graph-vs-knowledge-graph): How a context graph extends a knowledge graph with decision traces, provenance, time and policy for agents
- [Knowledge Base vs Knowledge Graph](https://flur.ee/blog/knowledge-base-vs-knowledge-graph): Documents vs entities and relationships, how each is queried, and how they work together for AI
- [Neo4j Limitations and What to Use Instead](https://flur.ee/blog/neo4j-limitations): Licensing, scale-out, performance and modeling limits of Neo4j, with sourced facts and alternatives
- [Context Layer vs Semantic Layer](https://flur.ee/compare/context-layer-vs-semantic-layer): What each layer defines, who uses it, and why most teams run both
- [Best Enterprise AI Search Software in 2026](https://flur.ee/compare/enterprise-ai-search): Glean, Coveo, Elastic, Microsoft 365 Copilot, Guru, Sinequa, and Fluree compared on connectors, permissions, RAG citations, agents, and pricing
- [Best Semantic Layer Tools for AI in 2026](https://flur.ee/compare/semantic-layer-tools): dbt, Looker, Cube, AtScale, Snowflake Semantic Views, Databricks Metric Views, and Fluree compared on definitions, agent interfaces, and pricing
- [Best Decision Intelligence Platforms in 2026](https://flur.ee/compare/decision-intelligence): Fluree, Palantir, Aera, Quantexa, SAS, and FICO compared on decision modeling, composite AI, orchestration, and governance
- [Fluree vs Neo4j](https://flur.ee/compare/fluree-vs-neo4j): Head to head: data model, Cypher support, time travel, access control, pricing, and migration
- [Fluree vs Stardog](https://flur.ee/compare/fluree-vs-stardog): Head to head: RDF, reasoning, access control, history, and licensing
- [Palantir Competitors and Alternatives in 2026](https://flur.ee/compare/palantir-alternatives): Alternatives split by the part of Foundry they replace, with sourced pricing evidence and when to stay on Palantir
- [Best Taxonomy Management Software in 2026](https://flur.ee/compare/taxonomy-management-software): Taxonomy and ontology tools compared on standards, governance, and publishing
- [Fluree vs Glean](https://flur.ee/compare/fluree-vs-glean): Head to head: the workplace-search index versus the governed knowledge graph, from permissions to pricing
- [Best Enterprise Knowledge Graph Platforms in 2026](https://flur.ee/blog/enterprise-kg-buyers-guide-2026): Neo4j, Memgraph, Stardog, Fluree, Graphwise, eccenca, and AWS Neptune compared against the same six criteria
- [Neo4j Alternatives in 2026](https://flur.ee/blog/neo4j-alternatives): Six graph databases compared on licensing, query languages, scale, and pricing, every claim sourced and dated
- [Glean Alternatives in 2026](https://flur.ee/blog/glean-alternatives): Seven enterprise AI search platforms compared on permissions, agents, deployment, and pricing, every claim sourced and dated
- [Weaviate Alternatives in 2026](https://flur.ee/blog/weaviate-alternatives): Seven vector search options compared on deployment, memory economics, hybrid search, and pricing

## Solutions

- [GraphRAG](https://flur.ee/solutions/graph-rag): Why knowledge-graph-grounded RAG beats vector-only retrieval
- [Semantic Layer](https://flur.ee/solutions/semantic-layer): A unified, queryable meaning layer across enterprise data
- [Conversational Analytics](https://flur.ee/solutions/conversational-analytics): Natural-language questions over governed data
- [Golden Records](https://flur.ee/solutions/golden-records): Entity resolution and master data management
- [AI Agent Governance](https://flur.ee/solutions/ai-agent-governance): Trusted data infrastructure for AI agents
- [Decision Intelligence](https://flur.ee/solutions/decision-intelligence): The decision intelligence platform built on a governed knowledge graph
- [Enterprise AI Search](https://flur.ee/solutions/enterprise-ai-search): Governed, cited answers across structured and unstructured enterprise data
- [What is AI-Ready Data?](https://flur.ee/solutions/ai-ready-data): Definitional guide: key characteristics, common challenges, analytics-ready vs AI-ready, and a readiness self-assessment
- [AI Data Preparation](https://flur.ee/solutions/ai-data-preparation): Guide to preparing unstructured content for AI: extraction, mapping, chunking vs knowledge graphs, linking prepared documents to structured tables, and choosing an approach
- [Entity Resolution](https://flur.ee/solutions/entity-resolution): Guide to entity resolution: how it works, challenges, terminology, and choosing software
- [Solutions by Industry](https://flur.ee/solutions/industries): Index of industry solutions: financial services, insurance, CPG, media, pharma, supply chain, and government
- [Financial Services](https://flur.ee/solutions/financial): Core banking AI, fintech operations, data monetization
- [Insurance](https://flur.ee/solutions/insurance): Policy, claims, and billing on one governed graph, with attribute-level access and as-of replay
- [Consumer Packaged Goods](https://flur.ee/solutions/cpg): Trade promotion, revenue growth management, product, and supply data in one governed graph for CPG teams and AI agents
- [Media](https://flur.ee/solutions/media): Content monetization and archival cataloguing
- [Pharma](https://flur.ee/solutions/pharma): Regulatory intelligence and clinical data intelligence
- [Government](https://flur.ee/solutions/government): Secure, verifiable data sharing for public sector
- [Supply Chain](https://flur.ee/solutions/supply): Traceability and multi-party data collaboration

## Demo Kits

Downloadable knowledge graphs an agent can load and verify without talking to sales. Each kit ships a standards-based model, instance data, and competency questions **with their expected answers**, so an agent can load the graph, answer the questions, and check its own work. Free to run: a hosted workspace at https://flur.ee/ai, or the CLI via `brew install fluree/tap/fluree`.

- [Demo Kits](https://flur.ee/demo): Catalog of loadable industry knowledge graphs, each with competency questions and expected answers
- [Consumer Goods Demo Kit](https://flur.ee/demo/cpg): Trade promotions, retailer agreements, and upstream traceability on the GS1 Web Vocabulary and GS1 EPCIS 2.0, over a real USDA FoodData Central corpus (CC0). 35 classes, 413 entities, 26 questions. Optional modules: time travel, governed access, decision memory
- [Financial Services Demo Kit](https://flur.ee/demo/financial): Commercial lending through a covenant breach (facilities, covenants, ratings, KYC), typed with FIBO business-entity ontologies over a real GLEIF LEI corpus (CC0). 19 classes, 171 entities, 18 questions. Optional module: governed access
- [Pharmaceutical Demo Kit](https://flur.ee/demo/pharma): An impurity investigation traced from API lot to batch to market recall, over a real openFDA drug listing corpus (public domain). 15 classes, 146 entities, 15 questions. Optional module: governed access

Each kit’s download is a zip containing `model.jsonld`, instance files, `shapes.json` (SHACL), `competency-questions.md`, and a `README.md` with load instructions. Fictional business facts never attach to the real licensed records in any kit; the model enforces the separation.

## Resources

- [Resources](https://flur.ee/resources): Whitepapers, reports, field guides, webinars, and videos
- [Blog](https://flur.ee/blog): Articles on GraphRAG, knowledge graphs, semantic data, and enterprise AI
- [Kevin Doubleday, Author](https://flur.ee/blog/author/kevin-doubleday): Communications Lead at Fluree; 45+ articles on knowledge graphs, semantic data, and enterprise AI
- [Brian Platz, Author](https://flur.ee/blog/author/brian-platz): Co-founder and Co-CEO of Fluree; writes on graph databases, verifiable data, and enterprise AI

## Company

- [About](https://flur.ee/about): Fluree’s mission, the four principles behind the product (trusted data first, built for the enterprise, open standards, a public benefit corporation), and office locations
- [Leadership](https://flur.ee/company/leadership): Executive team and advisory board, including co-founders Brian Platz and Andrew “Flip” Filipowski
- [Partner Program](https://flur.ee/company/partners): For SIs, ISVs, and platforms: build a context-layer practice, connect a data platform, or embed FlureeDB in a product, with enablement and joint go-to-market
- [Careers](https://flur.ee/careers): Open roles across engineering, product, and go-to-market
- [Contact](https://flur.ee/contact): Book a demo or talk to sales