eNRRCrew: Accelerating eNRR catalyst Design through Multi-Agent Collaboration and Automated Structure-Activity Analysis

The electrocatalytic nitrogen reduction reaction (eNRR) represents a promising approach for sustainable ammonia production. However, understanding structure-activity relationships remains challenging due to the vast literature and complex data analysis required. Here, we present eNRRCrew, a novel multi-agent collaborative framework that integrates large language models (LLMs), machine learning techniques, and automated data analysis tools to advance eNRR research. The eNRRCrew comprises five agents, an orchestrator, a yield predictor, a Faradaic efficiency predictor, a GraphRAG retriever, and a CSV file handler. Users interact with eNRRCrew through the user interface provided by the Streamlit library to perform retrieval and prediction of structure-activity relationships in eNRR.

• eNRR Yield predictor and FE predictor: - Using pre-trained machine learning models in the former section to predict eNRR yield and FE.
• GraphRAG retriever: - Enhanceing responses by retrieving information from curated databases containing eNRR abstracts.
• CSV file handler: - Writes and executes code to interact with CSV files obtained from text-mining workflow in response to user queries.

Online Demo

Try eNRRCrew on Online demo. Users can interact with eNRRCrew through an intuitive user interface provided by the Streamlit library, enabling them to efficiently retrieve and predict structure-activity relationships in eNRR.

Useful Links 🔗

• eNRRCrew demo vedio
• Microsoft's GraphRAG GraphRAG
• Microsoft's AutoGen AutoGen
• Steamlit Streamlit
• Microsoft's GraphRAG + AutoGen + Ollama + Chainlit = Fully Local & Free Multi-Agent RAG Superbot Medium.com 📚

📦 Installation and Setup

Follow these steps to set up and run eNRRCrew:

1. Create conda environment and install python packages:

   conda create -n eNRRCrew python=3.12.7
   conda activate eNRRCrew
   git clone https://github.com/nkuhuxu/eNRRCrew.git
   cd eNRRCrew
   pip install -r requirements.txt

2. Run eNRRCrew:

   streamlit run appUI.py