Mnemo AI – Local agentic assistant for any LLM that learns from its failures (github.com)

Mnemo AI has launched an innovative local agentic assistant designed to enhance interactions with various large language models (LLMs) through its Model Context Protocol (MCP) integration. Built using LangGraph and LangChain, this tool supports multiple providers like Ollama, Amazon Bedrock, OpenAI, and Anthropic, ensuring versatile usage across platforms. Users can quickly set it up with an interactive configurator that simplifies selecting models and features, promoting accessibility to advanced AI functionalities. The significance of Mnemo AI lies in its advanced features, such as automatic document indexing, personalized user profiles, and an episodic memory system that learns from both successes and failures, enhancing its conversational intelligence over time. Additionally, it integrates web search capabilities and efficient file handling, making it suitable for both casual users and developers working on complex tasks. Its support for vision models further broadens its utility, while the inclusion of smart error handling and background task management positions Mnemo AI as a robust tool ready to meet diverse AI needs. Licensed under the MIT License, it allows users to freely adapt the technology, fostering an environment of innovation within the AI/ML community.