AIROS

AI Research Operating System

Accelerating Scientific Discovery with AI Agents

Built for the Google × Kaggle AI Agents: Intensive Vibe Coding Capstone Project

🎥 Demo Video: ("https://youtu.be/s9KUQ1ruQag")

---

Overview

AIROS is an AI Research Operating System that transforms fragmented scientific literature into a structured, persistent research workspace.

Instead of functioning as a traditional chatbot, AIROS organizes research into dedicated workspaces where AI-powered pipelines assist with literature discovery, evidence extraction, knowledge organization, and intelligent research exploration.

---

The Problem

Modern research is scattered across:

• Academic papers
• PDFs
• Search engines
• AI chatbots
• Personal notes

Researchers spend significant time switching between tools, organizing information, and maintaining context throughout long-term projects.

---

The Solution

AIROS centralizes the research workflow into a unified workspace where researchers can:

• Create dedicated research projects
• Organize scientific literature
• Upload and analyze PDFs
• Explore structured evidence
• Visualize relationships through knowledge graphs
• Interact with an AI Research Copilot

---

Key Features

• 📚 Literature Workspace
• 📄 PDF Upload & Processing
• 🧠 Research Planner
• 🔍 Literature Retrieval
• 📊 Evidence Analysis
• 🌐 Interactive Knowledge Graph
• 🤖 Research Copilot
• 📂 Persistent Research Workspaces
• 🔒 Security Initialization
• 📈 Evaluation Dashboard

---

Research Workflow

Create Research Project
        │
        ▼
Research Planner
        │
        ▼
Literature Retrieval
        │
        ▼
PDF Processing
        │
        ▼
Evidence Extraction
        │
        ▼
Knowledge Graph
        │
        ▼
Research Copilot

---

System Architecture

---

Screenshots

Authentication

---

Dashboard

---

Create project

---

Research Planner

---

Literature Workspace

---

Evidence Workspace

---

Knowledge Graph

---

Research Copilot

---

Technology Stack

Frontend

• Next.js
• React
• TypeScript
• Tailwind CSS

Backend

• FastAPI
• Python

AI

• Google Gemini
• Google ADK
• Retrieval-Augmented Generation (RAG)

Database

• ChromaDB
• Supabase

---

Project Structure

AIROS/
│
├── frontend/
├── backend/
├── agents/
├── docker/
├── tests/
│
├── README.md
├── LICENSE
└── .env.example

---

Installation

Clone Repository

git clone https://github.com/YOUR_USERNAME/AIROS.git
cd AIROS

Install Frontend

cd frontend
npm install

Install Backend

cd backend
pip install -r requirements.txt

Environment Variables

Copy:

.env.example

to

.env

and configure your API keys.

---

Run AIROS

Frontend

cd frontend
npm run dev

Backend

cd backend
uvicorn app.main:app --reload

---

Google × Kaggle AI Agents Capstone

This project was developed as part of the Google × Kaggle AI Agents: Intensive Vibe Coding Capstone Project, demonstrating the application of AI agents to support scientific research workflows.

---

Future Scope

AIROS is designed with a modular architecture that can be extended with additional research capabilities while preserving a unified research workspace.

---

License

This project is licensed under the MIT License.

---

AIROS

Accelerating Scientific Discovery with AI Agents