You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A lightweight, reproducible demo of pharma competitive-intelligence graph construction and GraphRAG Q&A: 14 investor-deck PDFs → a Neo4j knowledge graph via neo4j-mcp-workspace-template, queried live in Claude Desktop. Includes data, extraction schema, demo skill, sample questions, and a pre-built DB dump.
An autonomous, 100% local Agentic GraphRAG system for codebase intelligence. Paste any GitHub URL or local directory to chat with your code. Powered by LangGraph, NetworkX, and Ollama (qwen2.5-coder).
AegisLogic (PCRA): A novel, local-first retrieval architecture (PCRA) for zero-hallucination Cyber Threat Intelligence. Synthesizes Graph (KuzuDB), Vector (Qdrant), and Temporal data via a LangGraph state machine. Includes empirical research on LLM evaluation metrics.
Moteur de connaissances local avec GraphRAG hybride (LanceDB + Kùzu). Il garantit la confidentialité grâce à un stockage sur site, tout en offrant une interface de chat bilingue et un explorateur 3D interactif.
End-to-end knowledge graph for Explainable AI research. Property graphs (Neo4j), RDF/OWL reasoning, KG embeddings (PyKEEN), and GraphRAG vs vanilla RAG benchmark on 3,907 arXiv papers.
A Temporal-Causal GraphRAG architecture for Advanced Persistent Threat (APT) Next-TTP prediction using actor-aware knowledge graphs and RotatE embeddings.
Eight RAG architectures answer the same query concurrently over one shared knowledge base, each scored by a common Trust Layer that adjudicates every answer claim-by-claim against its retrieved evidence — so you can see which retrieval strategy is actually more trustworthy, and which stage failed when it isn't.
An autonomous multi-agent DevSecOps pipeline that utilizes GraphRAG to detect vulnerabilities, map technical debt, and deploy surgical code patches directly via GitHub Pull Requests.