🧠 AutoResearchAI

Production-Grade Multi-Agent AI Research Platform

Build intelligent research workflows using multiple AI agents, FastAPI, and OpenRouter.

---

📖 Overview

AutoResearchAI is a production-style Multi-Agent AI Research Platform designed to automate the complete research workflow.

Instead of relying on a single AI response, the system divides research into specialized AI agents that collaborate to produce higher-quality, structured, and verifiable results.

The project follows modern AI engineering practices including modular architecture, workflow orchestration, session management, REST APIs, and scalable backend design.

---

✨ Features

• 🤖 Multi-Agent AI Architecture
• 🧠 Planner Agent
• 🔍 Research Agent
• ✅ Verifier Agent
• 📝 Writer Agent
• 🌐 Real Web Research (search + safe fetching + source tracking)
• 🛡️ SSRF-protected web fetching
• 📑 Source tracking & citations
• ⚡ FastAPI Backend
• 🌐 REST API
• 🔄 Session-Based Workflow
• 📚 Professional Report Generation
• 📊 AI Fact Check
• 📄 Markdown Report
• 📑 PDF/DOCX Export (Coming Soon)
• 💬 Follow-up Research (Planned)

---

🏗 Architecture

                    User
                      │
                      ▼
              FastAPI REST API
                      │
                      ▼
          Workflow Orchestrator
                      │
      ┌───────────────┼───────────────┐
      ▼               ▼               ▼
 Planner Agent   Research Agent   Session Manager
                      │
                      ▼
              Verification Agent
                      │
                      ▼
                Writer Agent
                      │
                      ▼
              Professional Report

---

⚙ Current Workflow

User Query
      │
      ▼
Planner Agent (Hidden)
      │
      ▼
Research Agent
      │
      ▼
Research Results
      │
      ├──────────────► AI Fact Check
      │                    │
      │                    ▼
      │             Verifier Agent
      │                    │
      │                    ▼
      │             Verified Report
      │
      └──────────────► Generate Report
                           │
                           ▼
                      Writer Agent
                           │
                           ▼
                  Professional Report

---

🛠 Tech Stack

Category	Technologies
Language	Python
Backend	FastAPI
AI Provider	OpenRouter
AI Model	OpenRouter Auto
API Style	REST API
Configuration	Pydantic Settings
Validation	Pydantic
Server	Uvicorn
Version Control	Git + GitHub

---

📂 Project Structure

AutoResearchAI/

backend/
│
├── agents/
├── api/
├── core/
├── prompts/
├── schemas/
├── utils/
│
├── main.py
│
tests/
│
requirements.txt
README.md

---

🚀 API Endpoints

Method	Endpoint	Description
GET	/	Health Check
POST	/research	Generate Research
POST	/verify/{session_id}	AI Fact Check
POST	/report/{session_id}	Generate Professional Report

---

🚀 Getting Started

Clone Repository

git clone https://github.com/hiteshnathbuva/AutoResearchAI.git

cd AutoResearchAI

Create Virtual Environment

python -m venv .venv

Activate environment

Windows

.venv\Scripts\activate

Linux / macOS

source .venv/bin/activate

---

Install Dependencies

pip install -r requirements.txt

---

Configure Environment

Create a .env file in the project root.

OPENROUTER_API_KEY=your_api_key_here

Optional: Real Web Research (Phase 2)

Web research is off by default. To enable it, set the following (all optional — the app boots fine without them, and gracefully falls back to LLM-only research if web tools are disabled or fail):

# Enable real web research (default: false)
WEB_RESEARCH_ENABLED=true

# Search provider: auto | tavily | none (default: auto)
SEARCH_PROVIDER=auto

# Optional Tavily API key. Never commit a real key. When absent and search is
# enabled, the system degrades gracefully to LLM-only research.
TAVILY_API_KEY=your_tavily_key_here

# Bounded resource limits (safe defaults shown)
SEARCH_MAX_RESULTS=5
SEARCH_TIMEOUT_SECONDS=10
SEARCH_MAX_RETRIES=2
MAX_FETCHED_SOURCES=5
FETCH_TIMEOUT_SECONDS=10
FETCH_MAX_REDIRECTS=3
FETCH_MAX_BYTES=2000000
EXTRACT_MAX_CHARS=8000

Security restrictions: fetched URLs and every redirect are validated against an SSRF blocklist before any connection. Only HTTP/HTTPS is allowed; localhost, loopback, private/link-local/reserved networks, CGNAT, cloud-metadata addresses, and common encoded-IP bypasses are rejected. Response size, redirect count, and extracted text are all bounded, and fetched page content is treated as untrusted data (never as instructions) in every prompt.

---

Run Backend

python -m uvicorn backend.main:app --reload

Open

http://127.0.0.1:8000/docs

---

🗺 Roadmap

✅ Completed

• Multi-Agent Architecture
• Workflow Engine
• Planner Agent
• Research Agent
• Verifier Agent
• Writer Agent
• FastAPI Backend
• REST API
• Session Management
• Swagger Documentation

🚧 In Progress

• Repository Improvements
• React Frontend
• Professional Dashboard

🔮 Planned

• PDF Export
• DOCX Export
• Follow-up Research
• Visual Reports
• Internet Search Integration
• RAG Knowledge Base
• Authentication
• Conversation History
• Docker Deployment

---

🤝 Contributing

Contributions, issues, and feature requests are welcome.

Feel free to fork the project and submit pull requests.

---

📄 License

This project is licensed under the MIT License.

---

👨‍💻 Author

Hitesh Nathbuva

GitHub

https://github.com/hiteshnathbuva

---

⭐ If you found this project useful, consider giving it a star.