🚀 ResearchFlow AI

Production-Grade Multi-Agent Research Assistant

FastAPI • LangGraph • LangChain • Groq • React • TypeScript

A modular AI research assistant that orchestrates multiple AI agents to generate structured, source-backed research reports.

---

📑 Table of Contents

• Overview
• Features
• System Architecture
• Backend Architecture
• LangGraph Workflow
• Backend Execution Flow
• Project Structure
• Tech Stack
• Installation
• API
• Future Improvements
• Resume Highlights

---

📌 Overview

ResearchFlow AI is a production-style multi-agent research assistant built using FastAPI, LangGraph, LangChain, Groq LLMs, React and TypeScript.

Instead of relying on a single LLM response, the application decomposes a research problem into multiple AI agents that collaborate through a shared LangGraph state.

---

✨ Features

• 🤖 Multi-Agent AI
• 🔄 LangGraph Stateful Workflow
• ⚡ FastAPI REST APIs
• 💻 React + TypeScript UI
• 📝 Markdown Report Generation
• 📚 Source Citation Support
• 🧩 Modular Architecture

---

🏗 System Architecture

User
 │
 ▼
React Frontend
 │
 ▼
FastAPI Backend
 │
 ▼
LangGraph Workflow
 │
 ├── Planner
 ├── Researcher
 ├── Verifier
 └── Writer
 │
 ▼
Markdown Report
 │
 ▼
JSON Response

---

🧠 LangGraph Workflow

flowchart TD
    A([START]) --> B[Planner]
    B --> C[Researcher]
    C --> D[Verifier]
    D --> E{Enough Information?}
    E -- No --> C
    E -- Yes --> F[Writer]
    F --> G([END])

Loading

---

⚙ Backend Execution Flow

sequenceDiagram
participant User
participant FastAPI
participant LangGraph
participant Planner
participant Researcher
participant Verifier
participant Writer

User->>FastAPI: POST /research
FastAPI->>LangGraph: Invoke Graph
LangGraph->>Planner: Plan
Planner->>Researcher: Collect Data
Researcher->>Verifier: Validate
Verifier-->>Researcher: Retry (if required)
Verifier->>Writer: Verified Findings
Writer->>FastAPI: Markdown Report
FastAPI-->>User: JSON Response

Loading

---

📁 Project Structure

backend/
├── app/
│   ├── api/
│   ├── agents/
│   ├── graph/
│   ├── core/
│   ├── services/
│   ├── schemas/
│   ├── utils/
│   └── main.py

frontend/
└── src/
    ├── components/
    ├── pages/
    ├── services/
    ├── hooks/
    ├── types/
    └── App.tsx

---

🛠 Tech Stack

Backend

• FastAPI
• LangGraph
• LangChain
• Groq
• Python
• Pydantic

Frontend

• React
• TypeScript
• Tailwind CSS
• Axios
• Vite

---

🚀 Installation

Backend

pip install -r requirements.txt
uvicorn app.main:app --reload

Frontend

npm install
npm run dev

---

📡 API

POST /research

{
  "query":"Explain LangGraph"
}

Response

{
  "report":"...",
  "sources":[]
}

---

🛣 Future Improvements

• PDF Upload
• RAG
• Docker
• Authentication
• Streaming
• Deployment
• History

---

💼 Resume Highlights

• Built a production-style Multi-Agent Research Assistant using FastAPI, LangGraph, LangChain, Groq and React.
• Designed a stateful LangGraph workflow with Planner, Researcher, Verifier and Writer agents.
• Developed scalable REST APIs with modular backend architecture.

---

📄 License

MIT

---

Made with ❤️ by Aditya Kumar