MAESTRO is an AI-powered research assistant designed for self-hosted, collaborative research. It automates complex tasks from planning to report generation, catering to academics, analysts, writers, and developers working with large datasets.

How It Works

MAESTRO employs a multi-agent framework, the WRITER Agentic Framework, where specialized AI agents (Controller, Planner, Researcher, Reflector, Writer) collaborate iteratively. This structured approach, involving research-reflection and writing-reflection loops, ensures transparency, critical review, and high-quality, evidence-based output.

Quick Start & Requirements

• Install: Clone the repository and run ./setup-env.sh for configuration, then docker compose up --build -d.
• Prerequisites: Docker, Docker Compose, Git, NVIDIA GPU (recommended), ~5GB disk space for models.
• Access: Web interface at configured address (e.g., http://localhost:3030), default login admin/adminpass123.
• Docs: DOCKER.md

Highlighted Details

• Advanced RAG pipeline optimized for academic and technical papers.
• Supports local, OpenAI-compatible LLMs and SearXNG for fully self-hosted operation.
• Iterative research-reflection and writing-reflection loops for quality assurance.
• Provides detailed transparency into AI agent reasoning and mission progress.

Maintenance & Community

• Actively maintained, accepting feedback and bug reports via GitHub Issues.

Licensing & Compatibility

• Dual-licensed: GNU Affero General Public License v3.0 (AGPLv3) and a separate commercial license available upon contact. AGPLv3 may have implications for commercial or closed-source use.

Limitations & Caveats

The AGPLv3 license requires that any modifications or derivative works distributed must also be made available under the same license, which could impact commercial integration.