getzep/graphiti
Graphiti
Graphiti is a backend framework and memory server designed to provide artificial intelligence agents with persistent, time-aware knowledge graph storage. It functions as a memory layer that enables agents to maintain context across long-term interactions by recording and evolving structured data over time.
The system distinguishes itself through a specialized temporal graph database that tracks how entities and relationships change using validity windows. By combining semantic vector similarity, keyword matching, and graph topology traversal, the engine performs hybrid retrieval to locate relevant information. It further refines these results by calculating graph distances from central entities, ensuring that retrieved context is prioritized based on its structural relevance to the query.
The platform supports schema-driven entity modeling, allowing for the enforcement of domain-specific structures on incoming data. It manages the ingestion of raw inputs into structured graphs and performs incremental updates to maintain the knowledge base without requiring full batch recomputation. Through standardized interfaces and protocol support, the system integrates with various large language model providers to automate data extraction and reasoning.
Features
- Agent Memory Engines - Provides persistent, time-aware memory storage for AI agents to maintain context across interactions.
- AI Memory Layers - Provides a persistent memory layer for artificial intelligence agents to maintain context across long-term interactions.
- Temporal - Ships a specialized storage engine that manages structured information with validity windows to support real-time updates and historical retrieval.
- Knowledge Graphs - Maintains temporal knowledge graphs to provide structured, time-aware context and long-term memory for agents.
- Model Context Protocol Servers - Exposes graph entities and relationships via the Model Context Protocol for real-time agent access.
- AI Agent Integrations - Exposes graph-based memory capabilities to external AI agents through a standardized server interface.
- Knowledge Graph Retrieval Systems - Retrieves relevant facts from temporal graphs by centering searches on specific entities.
- Graph Querying - Performs contextual search by calculating graph distances to prioritize relevant information for AI assistants.
- Hybrid Memory Stores - Orchestrates hybrid retrieval by combining semantic vector similarity, keyword matching, and graph topology traversal.
- Hybrid Search Engines - Implements a hybrid retrieval engine combining semantic, keyword, and graph-based techniques to extract relevant context.
- Knowledge Graph Construction Tools - Builds and maintains evolving temporal knowledge graphs using validity windows and incremental updates.
- Temporal Graph Versioning Systems - Tracks evolving entities and relationships using validity windows to maintain a time-aware history.
- Incremental Updates - Maintains knowledge bases through real-time incremental updates to avoid batch recomputation overhead.
- Persistent Chat Histories - Persists conversation episodes into a temporal graph to maintain long-term interaction context.
- LLM Provider Integrations - Provides unified configuration and connectivity for multiple large language model providers.
- Result Reranking - Refines query results by calculating graph distances from central entities to prioritize contextually relevant information.
- GraphRAG Frameworks - Real-time knowledge graph construction for AI agents.
- Retrieval Augmented Generation - System for building real-time knowledge graphs for agents.
- Schema-Driven Data Modeling - Enforces domain-specific structures on incoming data to ensure consistent knowledge graph categorization.
- Agent Client Protocols - Implements standardized communication protocols for external agents to query structured graph data.
- LLM Model Integrations - Integrates large language models with custom entity schemas to automate data extraction and reasoning.
- Model Integration Layers - Provides a unified configuration layer for integrating diverse local and hosted AI models.
- Entity Modeling - Allows definition of domain-specific entity types to structure knowledge representation.
- Semantic Information Retrieval - Retrieves graph information using a combination of semantic similarity and keyword-based matching to identify relevant facts.
- Temporal Data Ingestion - Ingests and processes raw data into time-based graphs by extracting entities and chronological facts.
- Search Strategy Configurations - Allows tuning and selection of search strategies to retrieve specific graph components based on optimized configurations.
