Skip to content

Welcome to Memoir's Documentation

Memoir is a high-performance semantic memory system for AI agents that brings Git-like version control to AI memory management. It replaces opaque vector databases with transparent, versioned, cryptographically secure memory storage using hierarchical semantic paths.

Key Features

  • Git-like Versioning: Branch, commit, merge, and rollback memories with cryptographic integrity
  • Semantic Paths: Replace UUID keys with meaningful paths like profile.professional.skills.python
  • O(log n) Lookups: Fast hierarchical search instead of expensive vector operations
  • Memory Aggregation: Automatic consolidation of related memories at semantic locations
  • Clean Architecture: Proper separation of storage, classification, and search layers
  • Multiple Search Engines: Choose between fast keyword-based or intelligent LLM-powered search

Quick Example

from memoir import ProllyTreeMemoryStoreManager
from memoir.classifier.intelligent import IntelligentClassifier
from memoir.search.intelligent import IntelligentSearchEngine

# Initialize components with dependency injection
store = ProllyTreeStore(path="./memory_store")
classifier = IntelligentClassifier(llm=llm)
search_engine = IntelligentSearchEngine(llm=llm, store=store)

# Create memory manager
memory_manager = ProllyTreeMemoryStoreManager(
    prolly_store=store,
    classifier=classifier,
    search_engine=search_engine
)

# Store memories with automatic classification
await memory_manager.store_memory(
    content="I work as a senior software engineer at TechCorp",
    namespace="user123",
    auto_classify=True
)

# Search with intelligent path selection
results = await memory_manager.search_memories(
    query="What is the user's job?",
    namespace="user123"
)

Performance

  • Search latency: 0.1-1ms (vs 150-750ms for vector databases)
  • Storage latency: 20-30ms (vs 200-600ms traditional)
  • Classification: 1-5ms pattern matching (vs 2-5s LLM-only)
  • Overall improvement: 10-20x faster end-to-end