MallMind is a state-of-the-art mobile-first Progressive Web Application (PWA) designed to revolutionize the physical mall experience through Agentic AI. It bridges the gap between traditional retail and digital intelligence, providing shoppers with a real-time, AI-powered concierge that handles everything from product discovery to complex indoor navigation.
In an era dominated by e-commerce, MallMind brings personalization and digital visibility back to the mall, allowing shoppers to interact with their environment conversationally and efficiently.
| AI Chat Concierge | Smart Navigation |
|---|---|
 |
 |
| Conversational AI understands intent and budget | Real-time route optimization between stores |
| Live Inventory Scout | Virtual Try-On (GenAI) |
|---|---|
 |
 |
| Simulated real-time stock and pricing checks | Advanced AI-powered visual try-on experience |
Physical malls currently suffer from:
- Fragmentation: Lack of personalized store recommendations.
- Invisibility: No real-time inventory visibility across brands.
- Staticity: Outdated static maps and zero route optimization.
- Lost Opportunities: Inability to leverage session-level visitor data.
We built a scalable PWA that:
- Deciphers Intent: Uses Natural Language Processing to understand shopping goals.
- Aggregates Inventory: Recommends relevant stores based on live availability.
- Optimizes Movement: Generates the most efficient walking routes.
- Visualizes Purchases: Enables AI-powered virtual try-on experiences.
MallMind is powered by a Multi-Agent Orchestration Layer. Instead of a single chatbot, a team of specialized agents collaborates to serve the user:
- 🕵️♂️ Inventory Scout: Queries store APIs for specific items, sizes, and pricing.
- 👗 Personal Stylist: Matches recommendations to the user's style and occasion.
- 🗺️ Smart Navigator: Calculates the shortest path between store centers using a mall-wide graph.
- 🏷️ Deal Negotiator: Identifies active promotions and offers tailored to the user.
MallMind/
├── frontend/ # Next.js 14 + React Application
│ ├── pages/ # App routes (Chat, Map, Onboard)
│ ├── components/ # UI Components (ChatBox, StoreCard, MapModal)
│ ├── styles/ # Global & Glassmorphism CSS
│ └── public/ # Static assets & icons
├── backend/ # Node.js + GraphQL API
│ ├── agents/ # 🤖 Multi-Agent Logic (Stylist, Scout, etc.)
│ ├── graphql/ # 🖇️ Schema & Resolvers (Apollo Server)
│ ├── models/ # 🗄️ MongoDB/Mongoose Schemas
│ ├── server.js # 🚀 API entry point
│ └── seed.js # 🌱 Database seeding script
└── photos/ # 🖼️ Application screenshots & assets| Stakeholder | Value Delivered |
|---|---|
| Shoppers | Faster, personalized, stress-free shopping experience |
| Store Managers | Increased footfall, higher engagement, and conversion insights |
| Mall Admin | Granular visitor analytics and operational efficiency |
- ≥30% reduction in total shopping time.
- ≥40% improvement in store recommendation relevance.
- ≥95% accuracy in indoor navigation and routing.
- Session Security: JWT-based authentication for temporary sessions.
- Data Encryption: HTTPS encrypted communication for all API traffic.
- Privacy First: No permanent storage of user photos used for virtual try-on.
- Auto-Expiry: Sessions automatically expire after inactivity to protect user data.
Scan the QR code below to instantly launch the MallMind concierge on your smartphone. Experience the full PWA features, including indoor navigation and AI chat, without any installation.
- Node.js (v18+)
- MongoDB Atlas Account
- OpenAI API Key
cd backend
npm install
# Configure .env with MONGO_URI and OPENAI_API_KEY
npm startcd frontend
npm install
# Configure NEXT_PUBLIC_API_URL in .env
npm run devMade with ❤️ by Prerita Saini.