LangR-multi-agent

This is reporter agent built with LangGraph. It uses wikipedia, arxiv and custom document retriever tools to complete tasks. Additionally, LangSmith is used to monitor and debug every user interaction.

Features

• Tools: It makes use of Wikipedia and Arxiv for research and gathering information.
• RAG System: Combines retrieval from document embeddings and OpenAI's LLM to provide context-specific answers.
• LangSmith Monitoring: Tracks and analyzes all interactions to improve the chatbot's performance and reliability.

Prerequisites

Before running the application, ensure you have the following:

1. Python 3.8 or later
2. Required Python libraries:
   • langGraph
   • Wikipedia
   • Arxiv
   • langchain-community
   • faiss-cpu
   • langchain
   • langchain-openai
   • langsmith

Folder Structure

• app.py: Main application file.
• tools.py: Defining tools for the agent

How to Run

1. Clone the repository:

   git clone https://github.com/BertrandConxy/langR-multi-agent.git
   cd langR-multi-agent

2. Create virtual env

   python -m venv venv

3. Install the dependencies using:

   pip install -r requirements.txt

4. Create .env file for credentials

   OPENAI_API_KEY
   LANGSMITH_TRACING=True
   LANGSMITH_ENDPOINT
   LANGSMITH_API_KEY
   LANGSMITH_PROJECT
   

5. Run the app:

   python app.py

Demo

What's Next

• Persisting custom knowledge documents locallly so that it doesn't take long for embeddings.

Monitoring with LangSmith

LangSmith is integrated into this project to monitor and analyze chatbot interactions. This ensures the app remains robust and user-friendly. To configure LangSmith:

1. Set up your LangSmith account and API key.
2. Ensure the LANGSMITH_API_KEY is added to your environment variables.