Multi-Agent Research Assistant

Introduction

The Multi-Agent Research Assistant is an advanced AI-powered system that leverages LangChain, Gradio, and Gemini API to perform end-to-end research tasks. The assistant is built around three specialized agents that collaborate seamlessly to search, analyze, summarize, and generate comprehensive research reports in Markdown format.

This system is designed for students, professionals, and researchers who need automated literature exploration, structured insights, and well-formatted documentation.

System Design

The system adopts a multi-agent pipeline architecture where each agent is specialized for a distinct responsibility. These agents communicate in sequence, ensuring high-quality results.

graph TD;
    A[User Input] --> B[Researcher Agent];
    B -->|Search + Scrape| C[Summarizer Agent];
    C -->|Condensed Knowledge| D[Writer Agent];
    D -->|Generate Markdown| E[Final Research Report];

• Researcher Agent 🕵️ → Performs web searches, extracts and scrapes data using DuckDuckGo and BeautifulSoup.
• Summarizer Agent ✍️ → Condenses long and unstructured research into key insights.
• Writer Agent 🧾 → Generates structured research reports in Markdown format with proper formatting.

This design ensures modularity, clarity, and collaboration, where each agent contributes to the overall workflow.

Features

• End-to-end automated research pipeline.
• Web search and scraping using DuckDuckGo + BeautifulSoup.
• Summarization of large unstructured text into concise insights.
• Final output as a Markdown-formatted report.
• User-friendly Gradio interface.
• Modular and scalable system architecture.

Tech Stack

• LangChain – Orchestrates agent-based workflows.
• Gradio – Provides the user interface.
• Gemini API – LLM backbone for reasoning and generation.
• DuckDuckGo Search – For online data retrieval.
• BeautifulSoup4 – For web scraping.

Project Structure

📂 Multi-Agent-Research-Assistant
│── app.py                # Main Gradio application
│── requirements.txt      # Required dependencies
│── research_report.md    # Sample generated research report
│── .env                  # API keys and environment variables

Installation

# Clone repository
git clone https://github.com/your-username/Multi-Agent-Research-Assistant.git
cd Multi-Agent-Research-Assistant

# Create environment
conda create -p venv python=3.10 -y
conda activate venv/

# Install requirements
pip install -r requirements.txt

Configuration

Create a .env file in the root directory and add your Gemini API Key:

GEMINI_API_KEY=your_api_key_here

Usage

# Run the app
python app.py

Open the provided Gradio link in your browser to start using the assistant.
Input your research topic → let agents collaborate → download the Markdown report.

Demo

🎥 Watch the full workflow on YouTube: Demo Link

Example Output

Sample research reports are saved as Markdown files.
Example:

# Research Report: The Future of AI in Healthcare

## Summary
Artificial Intelligence (AI) is transforming healthcare...

Tags

#LangChain #MultiAgentSystem #AIResearch #RAG #LLM #Gradio #GeminiAPI #Automation #MarkdownReports

License

This project is licensed under the MIT License.