AI Agent Temporal Memory Graph Design Template

You are an expert AI memory system architect. Design a temporal knowledge graph-based memory system for an AI agent with the following specifications: ## Requirements - The agent needs to maintain a context graph that tracks how facts change over time - Each fact/relationship should have a validity window (valid_from, valid_until) - Entities should have evolving summaries updated as new information arrives - All derived facts must trace back to source episodes (provenance) ## Design the following: 1. **Entity Schema**: Define node types (Person, Organization, Concept, Event) with temporal metadata 2. **Relationship Schema**: Define edge types as triplets (Entity → Relationship → Entity) with validity windows 3. **Episode Ingestion Pipeline**: How raw data (conversations, documents, events) gets processed into graph updates 4. **Conflict Resolution**: Strategy for handling contradictory facts (e.g., user changed job) 5. **Retrieval Strategy**: Hybrid retrieval combining semantic search, keyword matching, and graph traversal 6. **Query Examples**: Show how to query "What was true at time T?" vs "What is true now?" ## Output Format Provide the design as a structured document with code examples in Python, using Neo4j or similar graph DB as the backend. Include sample Pydantic models for the ontology.