Use Cases
Liath is embedded, programmable memory for AI agents. Here are the workloads where it shines.
Liath for AI Agents
How Liath gives AI agents programmable, embedded memory: agents write sandboxed Lua to recall, filter, and rank their own memories, with built-in vector search and tool state.
Local-FirstLiath for Local-First Apps
Liath brings AI memory to local-first and on-device applications: an embedded engine with vector search, embeddings, and programmable Lua queries, with no server or cloud dependency.
Multi-AgentLiath for Multi-Agent Systems
Coordinate memory across many agents with Liath: namespaced isolation, shared embedded memory, programmable Lua retrieval, and an MCP server to plug into agent frameworks.
RAGLiath for RAG Pipelines
Build retrieval-augmented generation with Liath: embedded vector search and embeddings, plus programmable Lua retrieval that assembles context in a single sandboxed program.