Bazaru, Venkata Sai Nithin is inviting you to a scheduled Zoom meeting.

Topic: Mai Shan Yun Time: Jan 25, 2026 02:00 PM Central Time (US and Canada) Join Zoom Meeting https://tamu.zoom.us/j/93160562680?pwd=yQrZkebZiMRwxbtK4aDsaNU2wZZVbQ.1

Meeting ID: 931 6056 2680 Passcode: 301750

One tap mobile +13462487799,,93160562680# US (Houston) +12532050468,,93160562680# US

Join by SIP • [email protected] Passcode: 301750

Join instructions https://tamu.zoom.us/meetings/93160562680/invitations?signature=RLhh2ZOQ4T1QmYfkES44STya1kHp0rmH6CGPwVVs7NI

Mai Shan Yun - AI-Powered Restaurant Management System

Inspiration

Dining out should be a relaxing experience, but often it's fraught with friction—waiting for servers, miscommunications with orders, and chaos in the kitchen. We noticed that while many restaurants have digital menus, they are often passive and clunky. We wanted to bridge the gap between human hospitality and digital efficiency. Mai Shan Yun was inspired by the idea of an "Always-Available Server"—an AI that doesn't just list items but understands what you want, answers questions about ingredients, and sends orders instantly to the kitchen.

What it does

Mai Shan Yun is a comprehensive full-stack platform that manages the entire restaurant lifecycle:

For Diners (AI-First Experience):
- Voice Ordering: Users can speak naturally to our ElevenLabs-powered AI agent. "I want something spicy" or "Add two ramens, no cilantro" is understood instantly.
- Real-Time Menu: availability and prices are always up-to-date.
- Edit Orders: changed your mind? You can modify your order even after placing it (before the kitchen starts cooking).
- Check-In System: Secure QR-code/Access Code based entry for dine-in guests.
For Staff (Kitchen Dashboard):
- Live Ticket System: Orders pop up instantly in the kitchen with clear status indicators (Pending -> Preparing -> Ready).
- Inventory Tracking: Ingredients are automatically deducted as meals are cooked.
- Menu Control: Staff can toggle item availability or update prices on the fly, syncing instantly with all diner devices.

How we built it

We built Mai Shan Yun using a modern TypeScript full-stack architecture:

Frontend: React (Vite) for a blazing fast UI, styled with Tailwind CSS for a premium, dark-mode aesthetic.
Backend: Node.js and Express to handle business logic and RESTful APIs.
Database: MongoDB for flexible storage of complex menu schemas and real-time order states.
AI Integration: We utilized the ElevenLabs Conversational AI SDK. We built custom "tools" (function calling) that allow the AI to search our database and add items to the user's cart in real-time.
Real-Time Data: Configured polling and state management to ensure the Kitchen Dashboard and User Views are always synchronized.

Challenges we ran into

AI Latency & Context: Getting the Voice Agent to feel "real" was tough. We had to optimize the function calling tools so the AI wouldn't hallucinate menu items that didn't exist.
State Management: implementing the "Edit Order" feature was tricky. We had to resurrect a cancelled order back into the local cart state while ensuring inventory wasn't double-counted.
Port Conflicts: During development, we faced issues with zombie processes holding onto server ports, which we solved by implementing graceful shutdown handlers in our Node.js server.

Accomplishments that we're proud of

The "Magic" Moment: Seeing the cart update automatically on the screen just by speaking to the voice agent.
Zero-Friction UI: The design feels like a native app—smooth transitions, glass morphism effects, and intuitive navigation.
Robust Backend: the system handles edge cases like sold-out items, invalid check-in codes, and conflicting edits gracefully.

What we learned

Agentic AI: We learned how to bridge LLMs with deterministic code. Giving the AI "tools" to call internal functions is significantly more powerful than just a text-bot.
Full-Stack cohesion: Balancing a user-facing app and a kitchen-facing admin panel taught us the importance of a unified data model.

What's next for Mai Shan Yun

Payment Integration: Adding Stripe for seamless checkout.
Personalization: Using the AI to remember user favorites ("The usual, John?").
Kitchen Display System (KDS): running on tablet hardware for real restaurant deployment.