{
  "id": "page:docs-supermemory-research-raw-04-introduction",
  "_kind": "Page",
  "_file": "wiki/docs/supermemory-research/raw/04-introduction.md",
  "_cluster": "wiki",
  "attributes": {
    "nodeKind": "Page",
    "sourcePath": "docs/supermemory-research/raw/04-introduction.md",
    "sourceKind": "repo-docs",
    "title": "Supermemory Introduction",
    "displayName": "Supermemory Introduction",
    "slug": "docs/supermemory-research/raw/04-introduction",
    "articlePath": "wiki/docs/supermemory-research/raw/04-introduction.md",
    "article": "\n# Supermemory Introduction\n\nSource: https://supermemory.ai/docs/introduction, https://supermemory.ai/docs/intro\n\n## Core Definition\n\nSupermemory is \"the Memory API for the AI era\" -- infrastructure for AI agent memory and context management. Achieves state-of-the-art performance on LongMemEval and LoCoMo benchmarks.\n\n## Key Characteristics\n\n- **Scalability** -- handles growing data volumes\n- **Performance** -- \"hyper fast\" operations\n- **Affordability** -- cost-effective pricing\n- **Production-Ready** -- suitable for real-world deployment\n\n## Main Components\n\n- **Memory APIs**: Composable APIs for memory operations and RAG\n- **User Profiles**: Contextual intelligence for LLMs combining static and dynamic facts\n- **SDK Integration**: Multiple SDKs for Python and TypeScript\n- **Connectors**: Real-time sync with Google Drive, Gmail, Notion, OneDrive, GitHub, web crawlers\n\n## Operational Flow\n\n1. **Input**: Users submit text, files, and chat conversations\n2. **Processing**: Supermemory indexes them and builds a semantic understanding graph tied to entities (users, documents, projects, organizations)\n3. **Retrieval**: At query time, the most contextually relevant information reaches the language model\n\n## Context Delivery Methods\n\n- **Memory API** extracts and maintains evolving user facts in real-time\n- **User Profiles** combine static baseline with dynamic episodic details\n- **RAG Integration** provides semantic search with metadata filtering and contextual chunking\n\nAll three share the same context pool when using identical user identifiers.\n",
    "documents": []
  },
  "outgoingEdges": [],
  "incomingEdges": [
    {
      "from": "page:docs-supermemory-research",
      "to": "page:docs-supermemory-research-raw-04-introduction",
      "kind": "contains_page"
    }
  ]
}