Fluree
A database that automatically connects your scattered documents, spreadsheets and files into a web of related facts, so an AI agent can actually understand how they relate instead of guessing.
🔗 Visit Fluree
Description
Most companies have their data spread across PDFs, spreadsheets, old databases, images and videos, with no easy way to tell an AI system how a customer record relates to a contract or a support ticket. Fluree takes that messy pile of files and automatically stitches it into a "knowledge graph" — a network of facts and relationships — so AI agents get accurate, connected context instead of hallucinating links that don't exist.
Fluree is a knowledge graph database and AI data platform built around an open-source core, FlureeDB (438 GitHub stars, Rust with Clojure/TypeScript components), plus an enterprise layer for governance, audit trails, and long-term agent memory. It ingests multiple data formats (PDFs, databases, CSVs, images, video), exposes an MCP-compatible integration for Claude, OpenAI, Gemini and Ollama agents, and is used by organizations including the U.S. Department of Defense, Morgan Stanley, the Associated Press and Warner Bros Discovery. The open-source database is free to start; the managed AI platform has a free tier, with enterprise pricing available on request.
💬 Our review
The short version: if you're trying to give an AI agent reliable long-term memory and accurate context across scattered enterprise data, Fluree's knowledge-graph approach with audit trails is built for exactly that, at a level of governance most startups won't need but big regulated organizations will.
💰 Pricing
📊 Global score
🤖 AI-enriched data
FlureeDB open source gratuit pour démarrer ; plateforme Fluree AI avec palier gratuit ; tarification entreprise sur devis.
Pros
Open source (438 étoiles GitHub) avec option entreprise au-dessus
Clients solides et vérifiables : DoD, Morgan Stanley, Associated Press, Warner Bros Discovery
Intégration MCP native avec Claude, OpenAI, Gemini, Ollama
Ingestion multi-format (PDF, bases de données, CSV, images, vidéo)
Cons
Positionnement clairement entreprise/gouvernance — probablement excessif pour un petit projet
Marché des bases de graphe déjà occupé par Neo4j, bien plus connu
Tarification entreprise opaque, à négocier au cas par cas
Date de fondation non communiquée