azcore - A professional multi-agent framework built on Python
To help you ship azcore-powered apps to production faster, check out our comprehensive agent orchestration and workflow management capabilities.
pip install azcoreazcore is the easiest way to start building hierarchical multi-agent systems and autonomous applications powered by LLMs. With under 10 lines of code, you can create sophisticated agent teams that collaborate, reason, and solve complex problems. azcore provides pre-built agent architectures, orchestration patterns, and workflow management to help you get started quickly and seamlessly incorporate intelligent agents into your applications.
We recommend you use azcore if you want to:
- Quickly build multi-agent systems with hierarchical coordination
- Implement advanced reasoning patterns like ReAct, Reflexion, and Self-Consistency
- Create autonomous agent teams with built-in collaboration and routing
- Build production-ready agent applications with persistence, caching, and monitoring
azcore supports multiple agent patterns and workflows including:
- 🎯 Agent Patterns: ReAct, Reflexion, Reasoning Duo, Self-Consistency
- 🌲 Workflow Types: Sequential, Concurrent, Hierarchical, Forest Swarm, Graph-based
- 🤝 Team Coordination: Agent routing, pattern matching, group chat, mixture of agents
- 🔄 Advanced Features: Reinforcement learning, state management, agent persistence
- Multiple Agent Architectures: Choose from ReAct, Reflexion, and custom agent patterns
- Hierarchical Organization: Build complex agent hierarchies with supervisors and coordinators
- Flexible Workflows: Sequential, concurrent, graph-based, and swarm workflows
- Agent Routing: Intelligent routing based on patterns, capabilities, and context
- State Management: Robust state tracking and persistence across agent interactions
- MCP Integration: Support for Model Context Protocol team building
- Reinforcement Learning: Built-in RL manager for agent optimization with synthetic data training
- Synthetic Data Generation: Train RL models without real user data for faster bootstrapping
- Caching & Performance: Smart caching for LLM calls and conversation history
- Production Ready: Comprehensive logging, error handling, and monitoring
For full documentation, see the API reference.
pip install azcoreazcore is released under the MIT License.
- GitHub: https://github.com/Azrienlabs/Az-Core
- Issues: https://github.com/Azrienlabs/Az-Core/issues
- PyPI: https://pypi.org/project/azcore/
Built with ❤️ by the Azrienlabs team.
Ready to build the next generation of AI agents? Start with azcore today!