RiskRate AI

Real-Time Health Insurance Cost Estimator

"Stop waiting for quotes. Get your health insurance estimate in 30 milliseconds, not 30 minutes."

🔴 Live Demo: RiskRate AI

RiskRate AI is a modern, high-performance dashboard that provides instant health insurance premium estimates. By bridging the gap between Data Science and Frontend Engineering, this project transforms a standard Python regression model into a responsive, interactive web application.

---

The Concept

Traditional insurance quotes require lengthy forms and backend processing. RiskRate AI flips the script:

• Zero Latency: The Machine Learning model is embedded directly into the frontend.
• Interactive UX: Prices update instantly as users drag sliders or toggle switches.
• Visual Feedback: Dynamic color-coding for BMI risk levels and smooth number animations.

Tech Stack

• Frontend: React 18 (via Vite) for high-performance rendering.
• Styling: Custom CSS Design System (No frameworks like Bootstrap/Tailwind).
  • Why? To demonstrate mastery of CSS variables, flexbox/grid layouts, and responsive design without relying on pre-built components.
• Machine Learning: Linear Regression (trained on the Medical Cost Personal Datasets).
• Architecture: "Edge AI" approach—weights extracted from Python and hardcoded into the client-side logic for O(1) inference time.

The "Secret Weapon" (Model Logic)

Instead of a heavy backend API, I reverse-engineered the regression model to run purely in the browser.

The Equation:

$$Cost = w_{age} \cdot Age + w_{bmi} \cdot BMI + w_{children} \cdot Children + w_{smoker} \cdot Smoker + b$$

The Weights (Derived from Python):

• Base Bias ($b$): -12,105.48
• Smoker Penalty: +$23,810.32 (The dominant factor)
• Age Factor: +$257.83 / year
• BMI Factor: +$321.94 / unit
• Children: +$473.69 / child

Features

1. Split-Screen Layout

A clean, modern interface separating inputs (Left) from results (Right), optimized for both desktop and mobile.

2. Smart Inputs

• Age Slider: Restricts input to insurable ages (18-65).
• BMI Visualizer: The slider track changes color (Blue → Green → Yellow → Red) to visually indicate health risk categories (Underweight to Obese).
• iOS-Style Toggles: Premium feel for binary choices like "Smoker".

3. Instant Feedback

The "Estimated Annual Premium" card isn't just static text. It features a custom AnimatedNumber component that interpolates values smoothly, giving the app a "living" feel.

Installation & Setup

1. Clone the repository

   git clone https://github.com/natinew77-creator/RiskRate-AI.git
   cd RiskRate-AI/risk-rate-ai

2. Install Dependencies

   npm install

3. Run Locally

   npm run dev

   Open http://localhost:5173 to view the app.

Author

Natneal B.

Disclaimer: This application is for educational purposes. The estimates are based on a public dataset and do not constitute real financial advice.