Mirix-AI/MIRIX

Mirix is a multi-agent personal assistant designed to track on-screen activities and answer user questions intelligently. By capturing real-time visual data and consolidating it into structured memories, Mirix transforms raw inputs into a rich knowledge base that adapts to your digital experiences.

What it solves

Mirix is a multi-agent personal assistant designed to build a comprehensive, long-term memory of a user's digital life. It solves the problem of AI assistants forgetting context and lacking a deep understanding of a user's specific habits, tools, and preferences by continuously observing screen activity and engaging in natural conversations.

How it works

The system uses a multi-agent architecture where six specialized memory agents manage different types of information: Core, Episodic, Semantic, Procedural, Resource, and Knowledge Vault. It captures visual data from the screen and processes multi-modal inputs (text, voice, images) to build structured memories. To maintain memory quality, it features an "Auto-Dream" consolidation process that reviews, merges duplicates, and resolves conflicts in stored data. Data is stored locally using PostgreSQL with BM25 full-text search and vector similarity support for efficient retrieval.

Who it’s for

Users who want a deeply personalized AI assistant that remembers everything they do on their computer and can retrieve specific information from past interactions or work processes.

Highlights

  • Multi-Agent Memory: Six distinct components for organizing different types of knowledge.
  • Screen Activity Tracking: Automatically converts visual screen data into structured memories.
  • Privacy-First: Long-term data is stored locally.
  • Auto-Dream Consolidation: An automated cleanup process to merge and refine memories.
  • Multi-Modal Support: Processes text, images, voice, and screen captures.

Related

  • Project
    peterfei/ai-agent-team

    A virtual AI development team with specialized roles (PM, Backend, Frontend, etc.) and a persistent memory system for managing complex software projects.

  • Project
    vellum-ai/vellum-assistant

    A personal AI assistant with eight types of memory that learns user behavior and takes proactive actions across multiple apps and communication channels.

  • Project
    redis/agent-memory-server

    A structured memory layer for AI agents that provides intelligent short-term session memory and persistent long-term context using a two-tier storage model.

  • Project
    RichmondAlake/memorizz

    A Python framework for building memory-augmented AI agents featuring multi-layered persistent memory, continual learning, and a control plane for orchestrating multiple agent workers.

  • Project
    beilusaiying/always-accompany

    A multi-purpose AI agent framework focused on context and attention mechanisms, providing layered memory and dynamic retrieval for long-term companionship, coding, and work.