> For the complete documentation index, see [llms.txt](https://leviaprotocol.gitbook.io/leviaprotocol/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://leviaprotocol.gitbook.io/leviaprotocol/core-concepts/memory-management/factual-memory.md).

# Factual memory

**Factual Memory** Fact memory stores discrete pieces of factual data that are extracted from conversations or interactions. These facts are agnostic of time or context and are stored in a **vector database** (e.g., Pinecone) for efficient retrieval. When integrated with the Levia protocol's social media learning capabilities, this memory system:

* Extracts and validates knowledge from social media platforms through NLP algorithms, cross-referencing with trusted sources
* Maintains metadata about information acquisition time and verification status
* Organizes facts hierarchically while preserving semantic relationships
* Updates continuously through social media monitoring, validating new information against existing knowledge
* Enables semantic search for contextually relevant information retrieval

This architecture allows Levia to maintain current knowledge while ensuring information accuracy.