This project is a proof of concept for running a local-first multi-agent system using: 🤖 Local LLMs via Ollama 🧩 Simple function/tool-call detection using <tool_call>... 🔍 Brave Search API or optional Brave MCP plugin server 🧠 Two collaborating agents: Searcher and Synthesizer
Overview
What is Multi-Agent Research POC?
Multi-Agent Research POC is a proof of concept for a local-first multi-agent system that utilizes local LLMs via Ollama and integrates with the Brave Search API to enable collaborative research between two agents: Searcher and Synthesizer.
How to use Multi-Agent Research POC?
To use this project, clone the repository, install the required dependencies, set up your Brave API key, run Ollama locally, and execute the main program to see the agents in action.
Key features of Multi-Agent Research POC?
- Local-first multi-agent architecture
- Integration with Brave Search API for live web research
- Tool-call detection for dynamic function execution
- Collaboration between Searcher and Synthesizer agents
Use cases of Multi-Agent Research POC?
- Conducting automated research on specific topics.
- Synthesizing information from multiple sources into coherent summaries.
- Enhancing research workflows with AI-driven insights.
FAQ from Multi-Agent Research POC?
- What is the purpose of the Searcher agent?
The Searcher agent performs live web searches to gather information based on user queries.
- How does the Synthesizer agent work?
The Synthesizer agent takes the output from the Searcher and generates a summarized response.
- Is there a graphical user interface?
Currently, the project does not include a GUI, but future improvements may add one.