Brain-inspired memory system for AI agents
episemic.remember()
and episemic.recall()
!
Get started with Episemic in just a few lines of code:
For non-async code:
These are the only classes and methods you need to know as a user:
The main async interface for storing and retrieving memories. Use this for modern async Python code.
View DocumentationSynchronous wrapper for traditional Python code that doesn't use async/await.
View DocumentationSimple memory object with properties like .text, .title, .tags, .created_at, and .metadata.
View DocumentationSearch result containing a Memory object and relevance score (0.0 to 1.0).
View Documentationawait episemic.remember(text, title=None, tags=None)
- Store a memoryawait episemic.recall(query, limit=10, tags=None)
- Search for memoriesawait episemic.get(memory_id)
- Get a specific memoryawait episemic.find_related(memory_id)
- Find related memoriesawait episemic.forget(memory_id)
- Delete a memoryawait episemic.health()
- Check system healthAdvanced documentation for developers who want to understand or extend the internal architecture:
Internal configuration management system for databases and services.
View DocumentationInternal Pydantic models for memory objects, search queries, and system state.
View DocumentationThe EpistemicAPI class that the simple interface uses internally.
View DocumentationMemory consolidation engine that transfers memories from fast to persistent storage.
View DocumentationMulti-path retrieval engine using vector similarity, tags, and graph traversal.
View Documentationpoetry install
from episemic import Episemic
await episemic.remember("Hello world!")
await episemic.recall("hello")