🚀 Kaggle Mentor AI

Your AI Companion for Kaggle Competitions

An AI-powered multi-agent platform that helps beginners and experienced data scientists navigate Kaggle competitions—from understanding datasets to generating notebooks and preparing submissions.

🏆 Kaggle Vibe Coding Agents Capstone Project

Empowering Kaggle enthusiasts with AI-driven guidance from dataset exploration to competition submission.

Kaggle Mentor AI was developed as part of the Kaggle Vibe Coding Agents Capstone Project 2026.

This project was created to explore the capabilities of modern AI agent systems in solving real-world machine learning workflow challenges. The goal was to design an intelligent assistant capable of guiding users throughout the entire Kaggle competition lifecycle—from dataset exploration to submission generation.

The project aligns closely with the competition's focus on:

Intelligent AI agents
Multi-agent collaboration
Developer productivity
AI-assisted workflows
Educational accessibility
Practical deployment of agent systems

I would like to thank the Kaggle team and the Google Developer ecosystem for organizing initiatives that encourage innovation in AI agents and human-AI collaboration.

🎯 Problem Statement

Kaggle competitions involve multiple complex stages including dataset understanding, exploratory data analysis, feature engineering, model selection, notebook creation, model evaluation, and submission generation.

Beginners often struggle to determine the next step, while experienced participants spend considerable time performing repetitive tasks.

Kaggle Mentor AI addresses this challenge through a collaborative multi-agent architecture where specialized AI agents work together to provide intelligent, contextual assistance at every stage of the competition lifecycle.

💡 Solution

Kaggle Mentor AI introduces a modular multi-agent system where each AI agent specializes in a specific task.

Instead of relying on a single general-purpose chatbot, multiple specialized agents collaborate to:

Analyze datasets
Perform exploratory data analysis
Recommend ML strategies
Generate starter notebooks
Guide submission preparation

📖 Overview

Kaggle Mentor AI is an intelligent multi-agent AI platform designed to simplify the entire Kaggle competition workflow.

Whether you're participating in your first Kaggle competition or aiming to climb the leaderboard, Kaggle Mentor AI provides structured, AI-powered assistance throughout your journey.

🤖 Google Agent Development Kit (Google ADK)

Kaggle Mentor AI follows a multi-agent orchestration architecture inspired by principles from the Google Agent Development Kit (Google ADK) ecosystem.

The system decomposes complex machine learning workflows into specialized AI agents, each responsible for a clearly defined task:

Agent	Responsibility
Dataset Agent	Dataset understanding and feature analysis
EDA Agent	Exploratory data analysis and visualization
Strategy Agent	Model recommendations and ML strategy
Notebook Agent	Starter notebook generation
Submission Agent	Submission preparation and competition guidance

This design follows the core Google ADK philosophy of:

Task specialization
Agent collaboration
Modular orchestration
Tool-augmented reasoning
Scalable AI systems

Rather than relying on a single monolithic AI assistant, Kaggle Mentor AI uses specialized agents that communicate through a prompt-routing layer to provide more accurate, explainable, and context-aware recommendations.

Google ADK Pattern Used

✅ Multi-Agent Architecture ✅ Agent Orchestration ✅ Tool-Augmented Agents ✅ Prompt Routing ✅ Specialized Agent Responsibilities ✅ Modular Agent Design

This architecture improves maintainability, scalability, and reasoning quality while enabling future expansion into persistent memory, autonomous workflows, and multi-LLM support.

🎥 Demo

YouTube Demo

https://www.youtube.com/watch?v=QY6ccoUa9B8

Live Demo

https://kaggle-mentor-ai-409406860748.asia-southeast1.run.app

✨ Features

📊 Dataset Analysis Agent

Automatic dataset understanding
Target variable detection
Missing value analysis
Feature overview
Data summary generation

📈 Exploratory Data Analysis Agent

Distribution plots
Correlation heatmaps
Missing value reports
Feature statistics
Dataset summaries

🤖 Strategy Agent

Provides recommendations for:

Data Cleaning
Feature Engineering
Model Selection
Cross Validation
Hyperparameter Tuning
Performance Optimization

📝 Notebook Generation Agent

Automatically generates starter notebooks containing:

Data Loading
Data Cleaning
Feature Engineering
Model Training
Evaluation
Prediction
Submission Creation

🏆 Competition Assistant

Helps users:

Understand competition objectives
Interpret evaluation metrics
Recommend workflows
Improve leaderboard performance

🧠 Multi-Agent Architecture

                    User
                      │
                      ▼
        React + TypeScript Frontend
                      │
             Prompt Routing Layer
                      │
      ┌──────────┬──────────┬───────────┐
      ▼          ▼          ▼           ▼
 Dataset     EDA Agent  Strategy   Notebook
  Agent                  Agent      Agent
      │
      ▼
 Submission Agent
      │
      ▼
 Gemini API

🧠 Why Multi-Agent?

Traditional Chatbot	Kaggle Mentor AI
Single Prompt	Multiple Specialized Agents
Generic Responses	Domain Expertise
Limited Context	End-to-End Competition Guidance
Difficult to Scale	Modular Architecture

🖥️ Tech Stack

Technology	Purpose
React	Frontend Framework
TypeScript	Type Safety
Vite	Build Tool
Tailwind CSS	Styling
Gemini API	AI Intelligence
Multi-Agent Architecture	Agent Collaboration

🚀 Installation

Prerequisites

Node.js v18+
npm
Git

Verify installation:

node -v
npm -v
git --version

Clone Repository

git clone https://github.com/Yuwin2008/Kaggle-Mentor-AI.git
cd Kaggle-Mentor-AI

Install Dependencies

npm install

Configure Environment Variables

Create .env

VITE_GEMINI_API_KEY=your_api_key

Run Development Server

npm run dev

Open:

http://localhost:5173

Production Build

npm run build

Preview Production Build

npm run preview

Lint

npm run lint

📜 Available Scripts

Command	Description
npm install	Install dependencies
npm run dev	Start development server
npm run build	Create production build
npm run preview	Preview production build
npm run lint	Run ESLint

⚙ Technical Highlights

Multi-agent architecture
Prompt routing system
Responsive UI
Type-safe codebase
Environment variable support
Production build support
Scalable frontend architecture

📸 Screenshots

Add screenshots for:

Home Page
Dataset Analysis
Auto EDA
Strategy Recommendations
Notebook Generation
Submission Assistant

🔄 Workflow

Competition
      │
      ▼
Dataset Analysis
      │
      ▼
Exploratory Data Analysis
      │
      ▼
Strategy Recommendation
      │
      ▼
Notebook Generation
      │
      ▼
Model Training
      │
      ▼
Submission File

🌟 Why Kaggle Mentor AI?

Beginner Friendly
AI-Powered Guidance
Multi-Agent Collaboration
End-to-End Workflow
Faster Learning
Intelligent Recommendations
Modern Responsive Interface

🔮 Future Roadmap

Authentication
Persistent Memory
AutoML
Team Collaboration
Cloud Notebook Execution
Leaderboard Analytics
Explainable AI
Dataset Version Tracking
Multi-LLM Support
Agent Memory

🤝 Contributing

Contributions are welcome.

Fork the repository
Create a feature branch
Commit your changes
Push the branch
Open a Pull Request

Name		Name	Last commit message	Last commit date
assets/.aistudio		assets/.aistudio
src		src
.env.example		.env.example
.gitignore		.gitignore
README.md		README.md
index.html		index.html
metadata.json		metadata.json
package-lock.json		package-lock.json
package.json		package.json
server.ts		server.ts
tsconfig.json		tsconfig.json
vite.config.ts		vite.config.ts

Latest commit

History

Folders and files

Repository files navigation