AmalGus — World's First B2B2C Glass & Allied Products Marketplace 💎

"India's Smartest Gateway for Architectural Glass Solutions"

AmalGus is a niche digital marketplace designed to bridge the structural gap in the $150–190 Billion global glass construction market. By connecting Homeowners, Architects, Builders, and Dealers directly with verified manufacturers and service partners, AmalGus brings transparency, speed, and AI-driven precision to a traditionally fragmented, WhatsApp-driven industry.

Built as a prototype in response to the AmalGus assignment brief by Preetam, Founder of AmalGus Technology.

---

🚀 Live Demo

https://amalgus-new.vercel.app

---

✅ Assignment Coverage

Core Requirements (4/4)

Feature	Implementation
Glass Product Catalog	Filter by type, thickness (mm), process & application. 10+ glass types with real industry attributes
Smart AI Matcher	Plain-language query → Groq LLM recommendation, with rule-based fallback for 8+ scenarios
Estimate / Quote Generator	mm → sq.ft conversion, GST (18%), role-based discounts (Builder 12%, Dealer 18%), allied product add-ons
Responsive UI	Role-aware design system, mobile-first layout, print-ready PDF estimate

Bonus Features (7/7)

Bonus	Implementation
Multi-vendor comparison	Sortable vendor table (price, rating, delivery days) on every product detail page
Daily Rate Dashboard	Live rate index with trend sparklines for 10 glass types
Allied Products	Cross-sell in estimator (UPVC Frame, Sealant, Hardware Fittings)
Service Partners	15+ verified partners, filterable by city & service type
Customer Role Selection	4 roles (Homeowner, Architect, Builder, Dealer) — each gets a personalized hero, catalog & CTA
Order History Mock-up	Full order history page with status filters, stats, and reorder flow
Allied products cross-sell	Paired products shown on product detail pages

---

🧠 AI Matching — How It Works

The Smart Matcher uses a two-layer architecture:

1. Groq LLM (primary): User's plain-language query is sent to llama3-8b-8192 with a system prompt that knows glass industry terminology, thickness norms, and application-based safety requirements.
2. Rule-based fallback (secondary): If the API is unavailable or rate-limited, a local scoring engine matches queries against 8+ predefined scenarios using keyword analysis, safety flags, and category priorities.

Example queries it handles:

• "I need glass for my bathroom shower" → Toughened, 8mm
• "Soundproof glass for office partition" → Acoustic Laminated, 8.8mm
• "External facade of a high-rise building" → Low-E / Reflective, 6mm+

---

🏗 Architecture

src/
├── app/
│   ├── page.js              # Role-aware homepage
│   ├── catalog/             # Product listing + detail pages
│   ├── smart-match/         # AI Matcher UI
│   ├── estimate/            # Quote generator with print support
│   ├── orders/              # Order history mock-up
│   ├── rates/               # Daily glass rate dashboard
│   ├── vendors/             # Factory partner directory
│   ├── service-partners/    # Installer / measurement listings
│   └── api/                 # Next.js Route Handlers (backend)
├── components/
│   ├── Navbar.js            # Role-aware responsive navigation
│   ├── RoleProvider.js      # Global role context (localStorage)
│   ├── ToastProvider.js     # Notification system
│   ├── VendorComparison.js  # Sortable vendor table
│   └── ...
└── data/
    ├── glassProducts.js     # 10+ glass types with full specs
    ├── vendors.js           # Mock vendor network
    ├── dailyRates.js        # Simulated market index data
    └── matchLogic.js        # AI + rule-based matching engine

---

💻 Getting Started

1. Clone the repository

git clone https://github.com/Vansh0204/amalgus-new.git
cd amalgus-new

2. Install dependencies

npm install

3. Environment Variables

Create a .env.local file:

NEXT_PUBLIC_GROQ_KEY=your_groq_api_key_here

Note: The app works without an API key — the rule-based fallback handles all matching scenarios.

4. Run Locally

npm run dev

Open http://localhost:3000

---

🛠 Tech Stack

Layer	Technology
Framework	Next.js 15 (App Router)
Styling	Tailwind CSS
AI	Groq API (llama3-8b-8192) with rule-based fallback
State Management	React Context API
Deployment	Vercel

---

🏭 Glass Industry Context

This prototype is built with real glass industry knowledge:

• Custom manufacturing — once cut, glass cannot be returned. The estimate tool reflects this with clear disclaimers.
• Daily rate volatility — the Daily Rates page simulates real market fluctuations.
• System-based selling — the estimator bundles glass with allied products (frames, sealants, hardware).
• Trust-based ecosystem — vendor ratings and verified badges reflect how B2B glass deals actually work.
• 52 customer types across 9 groups — simplified into 4 UI roles (Homeowner, Architect, Builder, Dealer).

---

Developed for the AmalGus Technology Internship Assignment — April 2026.
"We are building something that doesn't exist yet."