Multi-Agent Research Assistant System

AI Research Planner is a LangGraph-based multi-agent research workflow system that automates the full cycle of research — from topic planning and info gathering to synthesis, writing, fact checking, and critique — using specialized AI agents. It’s designed to be extensible, modular, and structured for iterative, conditional multi-agent execution. :contentReference[oaicite:1]{index=1}

---

Overview

Modern research tasks involve multiple steps like topic exploration, filtering facts, structuring insights, and producing quality writing. This project implements a multi-agent AI research pipeline where different agents — orchestrated with LangGraph — collaborate intelligently to:

• Research web and literature sources
• Analyze and synthesize findings
• Generate structured, human-readable output
• Apply checks and critiques for quality

Built with Python and web frontend components, this system demonstrates how autonomous, stateful agent workflows can accelerate complex research processes. :contentReference[oaicite:2]{index=2}

---

Features

Multi-Agent Architecture

• Planner Agent — sets goals, decides research direction
• Searcher Agent — locates relevant external info
• Analyzer Agent — summarizes and interprets data
• Writer Agent — creates structured reports
• Critic/Fact-Checker Agent — verifies accuracy

📌 State-Driven Workflow
Agents use a shared state model to manage decisions, routing, and iteration logic. :contentReference[oaicite:3]{index=3}

📌 Conditional Routing & Iteration Control
Workflows can adapt paths based on intermediate results and feedback from agents. :contentReference[oaicite:4]{index=4}

📌 Extensible and Modular
Designed to add new agents, tools, and integrations. :contentReference[oaicite:5]{index=5}

---

Getting Started

Prerequisites

Ensure you have the following installed:

• Python 3.10+
• pip or preferred Python package manager
• LangGraph and other dependencies (see below)

Installation

1. Clone the repository

   git clone https://github.com/YasithWijesuriya/Multi-Agent-Research-Assistant-System.git
   cd Multi-Agent-Research-Assistant-System
   

2. Install dependencies

    pip install -r requirements.txt
   

3. Configure environment variables (if any tools or APIs required)

4. Run the system

    python main.py
   

Project Structure

    Multi-Agent-Research-Assistant-System/
    ├── agents/                  # AI agent definitions
    ├── graph/                   # LangGraph workflow definitions
    ├── research-frontend/       # Frontend UI or client code
    ├── tools/                   # Tool wrappers and utilities
    ├── utils/                   # Shared helper functions
    ├── .gitignore
    ├── README.md
    ├── main.py                  # Main execution script
    ├── api.py                   # API entrypoints (if applicable)
    ├── pyproject.toml           # Python project config
    └── reports/                 # Sample output reports

How It Works

User provides a research query
Planner agent breaks down the task into subtasks
Searcher agent retrieves raw data (web, databases)
Analyzer agent cleans and interprets the data
Writer agent structures insights into outputs
Critic agent reviews the output for quality & consistency

Each agent updates a shared state model, and LangGraph handles the workflow and conditional transitions between tasks.