Antnest Chatbot

AI customer service connected to a dessert shop’s booking workflow

By · Updated

A LINE and web chatbot with a shared support workflow, reservations, calendar export, and an admin dashboard.

The product

Antnest Chatbot brings customer questions and reservations into one system for a dessert shop. Customers can use LINE or the web interface; the shop has a back office for products, settings, customers, and bookings.

How it works

The application combines keyword-based answers with Gemini-generated responses. LINE conversations retain context, and model failover provides another route when the primary model is unavailable.

Reservations support dates and flexible time slots, with an iCal export for calendars. A LIFF booking page works inside LINE and in a browser. The stack uses Next.js, TypeScript, and Supabase.

What this work demonstrates

The useful unit is the complete customer workflow: an answer, the relevant product information, and a way to make a reservation. The public repository documents these components and their integration.

Explore the project

About the developer

I am Stan Shih (施博瀚 / Po-Han Shih), an AI product developer and AI agent builder based in Taipei, Taiwan.

About Stan Shih 施博瀚 →