CoolFace
Apppublic

K2-k2/Multi-Agent-Research-Crew

sourceHugging Facemitupdated 6mo agoView on Hugging Face
0likes
App README

🚀 Multi-Agent Research Crew

A high-efficiency, memory-optimized AI research assistant built with Streamlit, CrewAI, and Google Gemini. This application orchestrates a collaborative team of AI agents to research complex topics and synthesize them into professional blog posts.

🌟 Overview

This project demonstrates how to leverage agentic workflows to solve multi-step problems. Instead of a single prompt-response interaction, this app utilizes two specialized agents:

  1. 1.Lead Researcher: Scours the topic to identify 5 critical, data-driven facts.
  2. 2.Tech Blogger: Transforms those facts into a polished, Markdown-formatted article.

🛠️ Key Features

  • —Secure API Management: Uses a sidebar input for your Gemini API key to keep it out of your source code.
  • —Memory Optimization: Configured with memory=False and allow_delegation=False to run efficiently on standard consumer hardware.
  • —Persistent Results: Utilizes Streamlit Session State to ensure your research results survive UI refreshes.
  • —Markdown Exports: Download your generated blog post instantly as a .md file.

💻 Local Installation & Setup

Follow these steps to get the Research Crew running on your machine:

  1. 1.Clone the Repository:
bash
   git clone [https://huggingface.co/spaces/K2-k2/Multi-Agent-Research-Crew](https://huggingface.co/spaces/K2-k2/Multi-Agent-Research-Crew)

2. Create a Virtual Environment (Recommended):

# Windows python -m venv venv venv\Scripts\activate

# macOS/Linux python3 -m venv venv source venv/bin/activate

  1. 1.Install Dependencies: Ensure you have a requirements.txt file, then run:
bash
   pip install streamlit crewai langchain-google-genai

4. Run the Application:

streamlit run app.py

🏗️ Architecture

The app follows a Sequential Process to ensure quality control:

AgentRoleResponsibility
ResearcherData DiscoveryFact-checking and trend identification.
WriterContent SynthesisNarrative structure and Markdown formatting.

1. Input: You provide a research topic and your Google Gemini API Key.

2. Execution: The Researcher agent finds the data, then passes it to the Writer agent.

3. Output: A formatted Markdown blog post is displayed and made available for download.

👨‍💻 About the Project

Developed by Kesar Agarwal an aspiring AI Engineer focused on creating accessible, agentic AI tools.

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference