🧠 Multi-Agent Research System

📑 Table of Contents

• 🧠 Multi-Agent Research System
  • 📑 Table of Contents
  • 🌟 Overview
  • 🚀 Getting Started
  • ✨ Features
  • 🗂️ Project Structure
    • Mermaid Diagram
    • File Tree
  • 🛠 Tech Stack
  • 🤝 Contributing
  • 👨‍💻 Author
  • 📄 License

🌟 Overview

The Multi-Agent Research System is a cutting-edge platform designed to revolutionize research by leveraging the power of multiple AI agents. This system allows users to input research topics, and then utilizes a team of specialized agents, orchestrated by LangChain and OpenRouter, to perform in-depth research, analysis, and report generation. The system is built with a modern web development stack, featuring a Next.js frontend and a FastAPI backend.

🚀 Getting Started

1. Clone the repository:

   git clone https://github.com/BjornMelin/multi-agent-research-system.git
   cd multi-agent-research-system

2. Install dependencies:

   • Make sure you have pnpm installed:

   npm install -g pnpm

   • Then, install project dependencies

   pnpm install

3. Run both frontend and backend:

   pnpm run frontend
   pnpm run backend

4. Open your browser and navigate to http://localhost:3000 for the frontend. The backend will typically run on port 8000.

✨ Features

• 🤖 Multi-Agent System: Utilizes multiple AI agents, each specialized in tasks like research, analysis, writing, and proofreading.
• 🧠 Advanced AI Models: Integrates with powerful language models, including DeepSeek-v3, LLaMA3.3-70B-Instruct, GPT-4o, and PHI-4 via OpenRouter.
• 📝 Research Topic Submission: User-friendly interface for submitting research topics and queries.
• 🔄 Real-time Progress Updates: Displays real-time updates on the progress of the multi-agent workflow using WebSockets or polling.
• 📊 Data Visualization: Provides a dashboard to visualize agent outputs and workflow stages.
• 📄 Report Generation: Automatically generates structured research reports based on agent findings.
• 📥 Report Download: Allows users to download reports in PDF or DOCX formats.
• 🗃️ Data Storage: Stores data locally using PostgreSQL and FAISS, with scalability for future AWS deployment.
• 🔍 Monitoring: Employs Prometheus and Grafana for system performance monitoring.

🗂️ Project Structure

Mermaid Diagram

graph TD
    A[User] --> B(Next.js Frontend);
    B --> C{FastAPI Backend};
    C --> D[Research Agent];
    C --> E[Analysis Agent];
    C --> F[Writing Agent];
    C --> G[Proofreading Agent];
    D --> H[DeepSeek-v3];
    E --> I[Python Libraries];
    F --> J[LLaMA3.3];
    G --> K[PHI-4];
    H --> L[OpenRouter];
    J --> L;
    K --> L;
    C --> M[PostgreSQL];
    C --> N[FAISS];
    C --> O[LocalStack];
    M --> P[User Queries, Workflow States];
    N --> Q[Document Embeddings];
    O --> R[PDFs, DOCXs];
    S[Prometheus] --> C;
    T[Grafana] --> S;

Loading

File Tree

multi-agent-research-system/
├── .gitignore
├── pnpm-lock.yaml
├── pnpm-workspace.yaml
├── package.json (Root)
├── backend/
│ ├── main.py
│ ├── requirements.txt
│ ├── venv/
│ └── ...
└── frontend/
├── app/
│ ├── about/
│ │ └── page.tsx
│ ├── api/
│ │ └── feedback/
│ │ └── route.ts
│ ├── error.tsx
│ ├── feedback/
│ │ └── page.tsx
│ ├── globals.css
│ ├── layout.tsx
│ └── page.tsx
├── components/
│ ├── forms/
│ │ ├── feedback-form.tsx
│ │ └── research-topic-form.tsx
│ ├── layout/
│ │ ├── footer.tsx
│ │ ├── header.tsx
│ │ └── success-message.tsx
│ ├── pages/
│ │ ├── about-page.tsx
│ │ ├── feedback-page.tsx
│ │ └── home-page.tsx
│ ├── sections/
│ │ ├── cta-section.tsx
│ │ ├── feature-card.tsx
│ │ └── hero-section.tsx
│ ├── theme-provider.tsx
│ └── ui/
│ ├── ... (ShadCN-UI components)
├── hooks/
│ └── use-toast.ts
├── lib/
│ ├── schemas.ts
│ └── utils.ts
├── public/
├── .eslintrc.json
├── components.json
├── next.config.js
├── package.json (Frontend)
├── postcss.config.js
├── tailwind.config.ts
└── tsconfig.json

🛠 Tech Stack

• Frontend:
  • Framework: Next.js
  • Language: TypeScript
  • Styling: Tailwind CSS
  • UI Components: shadcn-ui
  • State Management: React Context API or Zustand
• Backend:
  • Framework: FastAPI
  • Agent Orchestration: LangChain
  • Asynchronous Tasks: Celery or Ray
  • Web Server: Uvicorn
• Database:
  • Primary Database: PostgreSQL
  • ORM: SQLAlchemy
  • Vector Storage: FAISS (Local), Pinecone/Weaviate (Future)
• LLM Integration:
  • API Gateway: OpenRouter
  • Models:
    • DeepSeek-v3 (Retrieval, Summarization)
    • LLaMA3.3-70B-Instruct (Report Generation)
    • GPT-4o (Reasoning, Hypothesis Generation)
    • PHI-4 (Proofreading, Grammar)
• Monitoring:
  • Metrics: Prometheus
  • Visualization: Grafana
• File Storage:
  • Local: LocalStack (AWS S3 Emulation)
  • Cloud (Future): AWS S3
• Caching:
  • In-Memory Cache: Redis (Future)
• Package Manager: pnpm

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feat/version/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feat/version/AmazingFeature)
5. Open a Pull Request

👨‍💻 Author

Bjorn Melin

AWS-certified Solutions Architect, Developer, SysOps Administrator, and Machine Learning Engineer with expertise in cloud architecture and modern development practices. Connect with me on:

• GitHub
• LinkedIn
• Coursera
• Medium

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Built with ❤️ by Bjorn Melin

```