🧠 ExplainAI

Interpret machine learning models visually and interactively, directly in JavaScript.

🌟 Overview

ExplainAI is a production-ready JavaScript library that brings state-of-the-art model interpretability techniques to the web. Unlike Python-centric tools (SHAP, Captum, ELI5), ExplainAI runs client-side or edge-side, works with any model (local or remote), and integrates seamlessly into web applications.

Key Features

• 🔍 Multiple Explainability Methods: SHAP, LIME, Grad-CAM, Integrated Gradients, Attention Visualization
• 🌐 Universal Model Support: TensorFlow.js, ONNX.js, REST APIs, or custom implementations
• ⚡ High Performance: WebAssembly acceleration with pure JS fallback
• 🎨 Rich Visualizations: Interactive, customizable UI components
• 🔒 Privacy-First: Client-side processing, no data leaves your browser
• 📦 Modular Architecture: Use only what you need

📦 Packages

ExplainAI is organized as a monorepo with multiple packages:

• explainai - Complete package (includes all packages below)
• explainai-core - Core algorithms and model interfaces
• explainai-ui - React components and web visualizations
• explainai-node - Node.js runtime and CLI tools
• explainai-playground - Interactive demo application

🚀 Quick Start

Installation

# Install everything at once
npm install explainai

# Or install individual packages
npm install explainai-core
npm install explainai-ui

Basic Usage

import { explain } from 'explainai-core';
import { FeatureImportanceChart } from 'explainai-ui';

// Explain a model prediction
const explanation = await explain(model, input, { 
  method: 'shap',
  samples: 100 
});

// Visualize results
<FeatureImportanceChart explanation={explanation} />

📚 Documentation

• Getting Started Guide
• API Reference
• Examples
• Contributing Guide

🎯 Use Cases

• Model Dashboards: Embed explainability in monitoring tools
• AI Product UIs: Help users understand AI decisions
• Compliance Auditing: Provide visual proof for regulatory requirements
• Education: Teach interpretability in interactive environments
• MLOps Integration: Automated bias and interpretability checks

🏗️ Architecture

┌─────────────────────────────────────┐
│   Web UI / React Components         │
├─────────────────────────────────────┤
│   Explainability Engine             │
│   (SHAP, LIME, Grad-CAM, etc.)      │
├─────────────────────────────────────┤
│   Model Interface Layer             │
│   (TensorFlow.js, ONNX.js, API)     │
├─────────────────────────────────────┤
│   Computation Layer                 │
│   (WebAssembly + Web Workers)       │
└─────────────────────────────────────┘

🔧 Development

# Clone repository
git clone https://github.com/gyash1512/ExplainAI.git
cd ExplainAI

# Install dependencies
npm install

# Build all packages
npm run build

# Run playground
npm run dev

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

📄 License

MIT License - see LICENSE for details.

👤 Author

Yash Gupta (@gyash1512)

🙏 Acknowledgments

ExplainAI is inspired by excellent Python libraries:

• SHAP
• LIME
• Captum

---

Built with ❤️ for the JavaScript ML community