Memory
getzep/graphiti
Graphiti builds temporal knowledge graphs for agents, representing entities, relationships and how facts change over time. It is designed for continuously evolving knowledge rather than a static document index.
THE PRACTICAL EXPLANATION
Graphiti builds temporal knowledge graphs for agents, representing entities, relationships and how facts change over time. It is designed for continuously evolving knowledge rather than a static document index.
WHERE TO USE IT
Use it for relationship-aware memory, customer or project histories, operational intelligence and cases where the timing of a fact matters as much as the fact itself.
WHO MAY USE IT
Python data and agent teams that understand graph modelling, entity resolution and the consequences of storing derived knowledge.
HOW TO USE IT
Begin with a small domain and an explicit entity model. Feed dated events, inspect entity merging and query temporal changes. Build correction and provenance mechanisms before allowing the graph to influence important decisions.
GETTING THE BEST RESULTS
WHY IT MAY BE USEFUL
Build real-time knowledge graphs for agents.
Best considered for: Temporal knowledge and relationship-aware agent memory.
READ BEFORE YOU ADOPT IT
Graph quality depends on extraction accuracy and careful entity resolution.
Confirm the current licence, maintenance status, dependency risk, data path, model access, tool permissions and human approval points at the source. A public repository is inspectable raw material—not proof that a system is secure, supported or suitable for your production environment.