Project Alice is an open-source framework for creating, managing, and deploying AI agentic workflows. It targets developers and power users looking to build sophisticated AI applications by integrating task execution, intelligent chat, and flexible model deployment. The framework aims to provide a human-language interface for AI development, making it easier to engage with and control AI agents.

How It Works

Alice employs a microservices architecture with a Node.js backend, a Python workflow engine, and a React frontend. It supports a modular design where agents, tasks, and APIs interact through a defined data flow. Key features include a node-based task execution system, retrieval-augmented generation (RAG), human-in-the-loop mechanics, and support for various AI models (local and remote) and APIs. Data persistence is handled by MongoDB.

Quick Start & Requirements

• Installation: Clone the repository, create a .env file from the template, and run run.bat (Windows) or run.sh (Linux/Mac). Alternatively, use python run.py.
• Prerequisites: Git, Docker. LM Studio is optional for local models (recommended build 0.3.5 9-H).
• Setup: The process builds and launches containers. The frontend is accessible at http://localhost:4000/.
• Links: Installation videos: Part 1, Part 2. Live preview: ProjectAlice.

Highlighted Details

• RAG & Embeddings: Supports RAG with retrieval tasks and local embeddings, including semantic and recursive text splitting.
• Node Flow: Implements a node execution logic for tasks, allowing workflows to be composed of other tasks.
• Extensible APIs: Integrates numerous APIs including Google Search, Wolfram Alpha, PixArt Image Generation, and Bark TTS, with support for local and OpenAI-compatible LLMs.
• Data Clusters & Threads: Enables chats and tasks to hold updatable data clusters for context, and chats now use message threads for more organized conversations.

Maintenance & Community

The project is actively developed, with frequent updates listed in the README. Contributions are welcome via pull requests. For support, contact mampy3000@gmail.com.

Licensing & Compatibility

The source code is licensed under a BSD-style license. This license is generally permissive and compatible with commercial use and closed-source linking.

Limitations & Caveats

Project Alice is currently in alpha (v0.3.11), and users should expect changes. A database reinitialization is required for updates to v0.3. LM Studio integration requires a specific build (0.3.5 9-H). Some installation steps might require retries due to potential connection errors.