Warning
[PRE-Alpha] Please reach out if you want to get involved in the discussions. All feedback is welcomed
Bee Hive is a multi-agent platform designed to build, deploy, and serve powerful agentic workflows at scale. This repository contains implementations for both TypeScript and Python, enabling flexibility and interoperability for developers. The platform simplifies the creation and orchestration of AI agents powered by popular LLMs such as Llama 3.x and IBM Granite.
-
🔓 Open Platform: Run and orchestrate agents seamlessly on an open platform designed for experimentation and scalability.
-
🤖 Flexible Agent Framework: Write and experiment with agent and tool building in a flexible and modular framework.
-
⛓️💥 Versatile Agent Orchestration Capabilities: Design multi-agent workflows tailored to your specific needs.
-
✨ Optimized for Open-Source Models: Achieve best-in-class performance against open-source models like IBM Granite and Llama 3.x
This repository includes the following key modules:
-
Bee Agent Framework
1.1 TypeScript
The Typescript implementation of the Bee Agent Framework is available at: https://github.com/i-am-bee/bee-agent-framework
1.2 Python
The Python implementation of the Bee Agent Framework is available at: https://github.com/i-am-bee/bee-hive/tree/main/framework/bee-py
-
Bee Hive
The Bee Hive module serves as a multi-agent platform for orchestrating workflows. It supports declarative agent definitions and YAML-based workflows. Refer to the Bee Hive README for detailed instructions.
-
API 2.0: a powerful and lightweight API to help you integrate your agents
-
Dev UI: build and optimize your agentic workflows in a developer-friendly UI
-
Additional multi-agent workflows: conditionals, looping, async, meta-agent orchestration
-
Agent and tool library: shareable assets to help you get started
-
Streamlined experience: integrate key developer functionality in one repository
We welcome contributions from the community! Please check our CONTRIBUTING.md for guidelines and our Code of Conduct for standards of behavior.
Report bugs or feature requests via GitHub Issues. Ensure to check for duplicates before filing a new issue.
All content in this repository is provided by IBM under the associated open source software license. IBM is under no obligation to provide enhancements, updates, or support for this code.
Happy coding with Bee Hive 🚀