Cocktail Bot

Using our 'RiBot' scripting language, my team built a bartender chatbot that educates users on cocktail mixology while driving e-commerce. The bot connected to a cocktail delivery service.

Ecommerce User Journey


Using our RiBot scripting language, we can direct customers through a user journey from basic recommendations to the integrated product purchase page. By having a real scripting language, rather than just point and click flowcharts, we were able to build a scalable bot and update weekly with new features.

User Preferences

aj-tagging Based on the conversation we can understand user preferences and apply tags. These tags are synced with the wechat backend for content marketing.


aj-tagging Micro-tagging goes beyond just favorite drinks. We can remember a user and pick up the conversation across visits.



At any point a real human agent can step in and takeover the conversation. The users chat history with the bot, and tag information is available to the agent. We also provide shortcuts to 'hand back' to the bot, for example to ask a series of questions or drive to the store.

Support the support agents

livechat We still used the hybrid platform, but the bot took away the tedious work allowing live sales agents to focus on engaging customers with jokes, or upselling the cocktail delivery service.

View a video here!