Inspiration

My parents spent ₹45,000 on a washing machine that Pune's hard water destroyed in 6 months. The reviews never mentioned hard water. The YouTuber was from Delhi. The 4.2 stars meant nothing.

That one moment revealed a critical problem - Indian buyers face challenges no existing indian ecommerce online apps review site understands — voltage fluctuations, extreme heat, hard water, power cuts, joint families. ApplianceWise was built to fix this.

What it does

ApplianceWise is India's first AI-powered, one-stop appliance buying guide.

🤖 AI Recommendation Wizard Answer 4 questions (appliance, family size $n$, home type, budget $B$). Gemini AI returns your top 3 matches using:

$$S = 0.40R + 0.30P + 0.20B_{fit} + 0.10N$$

Where $R$ = rating, $P$ = positive sentiment, $B_{fit}$ = budget fit, $N$ = Indian needs match. 😊 Sentiment Analysis 10,000+ Indian reviews classified as positive / neutral / negative — surfacing real concerns like hard water damage, voltage issues, and summer heat.

💰 Live Price Comparison Real-time prices from all 4 major Indian retailers:

Best buy link along with fastest delivery dates

Store Network
🛒 Flipkart Pan-India
📦 Amazon.in Pan-India
🏪 Croma 200+ cities
💻 Reliance Digital 3,000+ stores

🔄 Side-by-Side Compare Compare up to 3 products. Best value highlighted. AI Verdict explains which to buy and why.

How we built it

Layer Technology
Framework Next.js 14 App Router
Styling Tailwind CSS + shadcn/ui
AI Engine Google Gemini 1.5 Flash
Price Data SerpApi $(gl=in,\ location=India)$
Deployment Vercel

Price caching with ISR:

const res = await fetch(url, { 
  next: { revalidate: 3600 }
});

Prices refresh hourly — keeping data fresh while staying within SerpApi's free tier.

Challenges we ran into

Brand images blocked hotlinking Sony, Samsung, LG CDNs block external image requests. Fixed by switching to Unsplash CDN with onError fallbacks showing the brand name.

ESLint failing Vercel builds 6 lint errors blocked production deployment despite working locally. Fixed with targeted .eslintrc.json rule overrides.

AI wizard showing wrong category Missing data.category in useEffect dependency array caused stale closure — recommending TVs for an AC search.

useEffect(() => {
  if (step === 4) {
    const filtered = products.filter(
      p => p.categorySlug === data.category
    );
    if (filtered.length === 0) { setStep(1); return; }
  }
}, [step, data.category]); // fix: added data.category

Buy links losing the user Store links were opening in the same tab. Audited every <a> tag and enforced:

target="_blank" rel="noopener noreferrer"

Accomplishments that we're proud of

  • 🇮🇳 First India-specific appliance AI — understands voltage, hard water, heat, and joint family needs that no US tool covers
  • 🤖 Gemini match score — personalised recommendations with transparent reasoning, not black-box suggestions
  • 💰 4-retailer price comparison — Flipkart, Amazon, Croma, Reliance Digital in one view
  • 😊 Sentiment at scale — 10,000+ Indian reviews classified and visualised
  • Ships as a real product — live on Vercel, open on GitHub, free for all Indians
  • 🆓 Zero paywalls, zero ads, zero bias — pure consumer value

What we learned

Technical

  • ISR revalidate caching is the right pattern for expensive external APIs at scale
  • React useEffect dependency arrays cause subtle but critical bugs when incomplete
  • Prompt engineering is architecture — Gemini's output quality is a direct function of how well Indian context is encoded in the prompt

Product

  • India is not one market — a recommendation for Mumbai differs from one for Rajasthan
  • Indian buyers mention service network quality far more than Western buyers — a uniquely Indian purchase concern that shaped our review analysis
  • Trust is the real product — consumers have information overload, not information shortage

What's next for ApplianceWise — India's AI-Powered Appliance Buying Guide

  • 🌐 Hindi and regional language support
  • 🔔 WhatsApp / email price drop alerts
  • ⭐ User-submitted verified reviews
  • 📱 PWA for offline browsing in low-connectivity areas
  • 🏘️ New categories: Geysers, Microwaves, Dishwashers, Air Purifiers, Vacuum Cleaners
  • 📊 30-day price history charts

🌐 Live: https://appliancewise-rht-raj.vercel.app 🐙 GitHub: https://github.com/rhtmundke-eng/appliancewise

Built With

  • bolt
  • chatgpt
  • claude
  • co-pilot
  • gemini
  • google
  • googlestudio
  • next.js
  • openrouterapi
  • tailwindcss
  • vercel
  • windsurf
Share this project:

Updates

posted an update

ApplianceWise — Built for Indian Families, Not Global Assumptions

The Problem

Every year, Indian families spend lakhs on the wrong appliance.

They open multiple browser tabs. They read fake reviews. They watch YouTubers from completely different climates and lifestyles.

Then they guess:

  • Will this AC survive a Nagpur summer?
  • Can this washing machine handle Chennai’s hard water?
  • Will the service network work in smaller cities?

And six months later… regret starts.

I built ApplianceWise to end that guesswork.


AI That Understands India

Most recommendation engines are built for Western markets.

ApplianceWise uses Gemini AI to generate personalized recommendations based on:

  • Family size
  • Home type
  • Budget
  • Indian usage conditions

Users instantly get:

Top 3 appliance matches

  1. AI match score
  2. Plain-English reasoning
  3. India-specific insights

Real Sentiment, Not Just Star Ratings

A 4.3 rating doesn’t tell you:

  • if the product survives voltage fluctuations
  • whether cooling drops in extreme summers
  • or if after-sales service is terrible

So we analyzed 10,000+ Indian buyer reviews and classified them into:

Positive Neutral Negative

This surfaces what Indian consumers actually love — and complain about.


Best Price. One Screen.

No more switching between tabs.

ApplianceWise compares live prices across:

  • Flipkart
  • Amazon
  • Croma
  • Reliance Digital

Users instantly see the best available deal with side-by-side comparison.


Built for India

Indian homes have challenges global review platforms rarely understand:

  • Hard water
  • Extreme heat
  • Voltage fluctuations
  • Power cuts
  • Joint family usage patterns

ApplianceWise is designed specifically around these realities.


The Vision

“ApplianceWise is what happens when every Indian family gets a brilliant, unbiased friend who has already done all the research — completely free.”


Why This Matters

Impact Value
Indian appliance market ₹2L Cr yearly
Time saved per purchase 3–5 hours
Cost to users Completely free
Ads / Sponsored bias None

Built Solo with AI

This project was independently conceptualized and built using:

  • Claude
  • Gemini AI
  • GitHub Copilot
  • Cursor
  • Bolt
  • YouTube learning
  • Vibe coding + relentless debugging

A product idea inspired by real Indian household problems — transformed into a working AI-powered platform.

Log in or sign up for Devpost to join the conversation.