Skip to main content

AI for us: The memory system for your AI agents

Project description

_

AIUS Logo

License: MIT GitHub last commit Version AIUS Discord AIUS Twitter


● AIUS ●

⚡ The long-term memory system for your AI, powering ... ⚡

  • 🥷 Superagents and 🤖 AI Agents
  • 🗣️ AI assistants
  • 🧠 LLMs

Built with GraphRAG, knowledge graphs, autonomous ai agents, latest research and a bit of magic ✨


🌟 Why AIUS - the big cause

We aim to allow AI assistants and Superagents to remember and understand:

  • You & Other Humans - think: Empathetic AI assistants engaged in Conversational AI with you and other individuals they interact with.
    • ≠ Not just ≠ limited context-windows, system prompts, pre-trained/fine-tuned knowledge or RAG that doesn't work.
      • ✔️ Learning you, from you and for you. Everything on-the-fly.
        • 🤖 📚 💡 👨 👩
  • Themselves - give them: self-consciousness. They'll know why they exist, their values, goals, preferences, memories of past interactions, tools and relationships.
    • ≠ Not just ≠ system prompts, prompt-engineered agent definitions with name, role, objectives, and function callings.
      • ✔️ Having self-awareness that is: autonomous, dynamic, programmable, adaptable, controllable and safe.
        • 🥷 🧠 🕰️ 🗣️ 🔒
  • Content - envision: Specialized agents being able to remember and understand any content they are exposed to.
    • ≠ Not just ≠ static files from private data corpora in limited file formats
      • ✔️ Processing multi-modal first-seen content in any format, streamed & static, in any language they know.
        • ➡️ 💬 🎵 🎬 📄 📡
  • Other agents and assistants - imagine the AInet - an Internet of Agents, a new ecosystem of AI assistants, AI agents, humans and organizations of these.
    • ≠ Not just ≠ pre-trained, piped agents with Human-in-the-loop (HITL)
      • ✔️ Building lasting relationships with other assistants, agents, humans and organizations
        • 🌐 🗣️ 🤖 👥 💻 💼

It's a promise that keeps us going 👣 and shipping ⛴️ at a lightning speed ⚡️


💡 Key features & design patterns

  • A configurable MemorySystem composed of a MemoryStorage layer, MemoryTypes, and MemoryPods.
  • MemoryPods: configurable Pods that allow setting up separate storage units for different memory types, providing a security isolation layer.
  • Custom MemoryStorages: think KV, Graph, Vector, and other DBMSes under the hood of MemoryPods.
  • MemoryTypes: support for multiple memory types, i.e., episodic memory, entity memory, working memory (WM), short-term (STM), and long-term memory (LTM), etc.
  • Simple high-level ontology:
    • Beings
      • Events (relating to energy)
      • Entities (relating to matter)
      • Relationships (relating to transformations)
  • A composable system that allows:
    • Building beings—think agents, assistants, robots—and commissioning them with:
      • Input interfaces (Sensors)
      • A memory system
      • Processing functions
      • Output interfaces (Tools)

We start small, ship, listen to feedback, break things, launch and go again.


🚀 Getting Started

# Install the repository
pip install aius

🔗 Join the community


🤝 Contribute

  • ⭐️ Star the repo and become one of our Stargazers
  • 💬 Join our discord to chat with the growing community about designs, contribution and use cases
  • Clone: git clone git@github.com:markmbain/aius.git and play around
  • 🍴 Fork, develop code, test, document, commit, create pull requests
  • Create New issues
# Clone the repo
git clone git@github.com:markmbain/aius.git
cd aius

# Install dependencies
pip install poetry
poetry install --with dev

📝 License

MIT License, Copyright (c) 2024 Mark Bain

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

aius-0.0.0.tar.gz (4.4 kB view details)

Uploaded Source

Built Distribution

aius-0.0.0-py3-none-any.whl (6.3 kB view details)

Uploaded Python 3

File details

Details for the file aius-0.0.0.tar.gz.

File metadata

  • Download URL: aius-0.0.0.tar.gz
  • Upload date:
  • Size: 4.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.7.1 CPython/3.9.19 Linux/5.15.0-52-generic

File hashes

Hashes for aius-0.0.0.tar.gz
Algorithm Hash digest
SHA256 47aa185a63310628911567081f67fe1e76c583e910a08dc6540f83e0d9879dc0
MD5 f5065fbec82bc0a474c05fd061c00125
BLAKE2b-256 c35f4011464fe0d52f1f8ffe6d3fb9ca655f2707fae4eb545195e36673201991

See more details on using hashes here.

File details

Details for the file aius-0.0.0-py3-none-any.whl.

File metadata

  • Download URL: aius-0.0.0-py3-none-any.whl
  • Upload date:
  • Size: 6.3 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.7.1 CPython/3.9.19 Linux/5.15.0-52-generic

File hashes

Hashes for aius-0.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 6e531a119301f2fcb6e300d5d0b6894a3af08c30a37f0cd96977b93c4b1d7205
MD5 1c36f9603a5f36b1f4de7c8d3e932924
BLAKE2b-256 f268a30cc64fd2924f3916748b783b2d0fd4cb08a0a0b6cfbee766fc8ffa1823

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page