Show HN: Mnemora – Serverless memory DB for AI agents (no LLM in your CRUD path) (github.com)

Mnemora has announced an open-source, serverless memory database specifically designed for AI agents, featuring four distinct memory types—including working, semantic, episodic, and procedural—accessible via a single API. Notably, Mnemora enables operations without the need for large language models (LLMs) in the CRUD path, achieving impressive sub-10ms read speeds and streamlined performance. Users can easily store and search memory states, which are auto-embedded as vectors for efficient retrieval. This innovation is significant for the AI/ML community as it simplifies memory management for AI agents, reducing reliance on LLMs and thereby lowering operational costs and complexity. With built-in multi-tenancy and scalable architecture, Mnemora fully utilizes serverless infrastructure, including DynamoDB for working memory and Aurora for semantic storage. The system is not only self-hostable but seamlessly integrates with popular frameworks like LangChain and CrewAI, making it a versatile tool for developers looking to enhance AI agent capabilities.