Restaurant-Ai-Agent

The Restaurant-Ai-Agent is a multi-stage automation system built in n8n that connects Google Sheets and Telegram with an AI agent at its core. It is designed to handle restaurant ordering through a conversational AI that talks to customers, understands their order in plain language, collects necessary details, and writes the order directly into the restaurant's order log. The system starts when a customer sends a message to the restaurant's Telegram bot, triggering the flow. The AI agent, powered by an OpenRouter-connected chat model, uses a system prompt containing the full menu and behavioral rules to reply in the customer's language, ask clarifying questions, and only mark an order as chosen once a specific item and the customer's name and phone number are provided. The agent is wired to three sub-components: a model for reasoning, a Window Buffer Memory node keyed to each customer's Telegram chat ID for conversation memory, and a Structured Output Parser that forces responses into a strict JSON schema. The agent's reply is sent back to the customer on Telegram, and a Switch node branches the flow based on whether the order is finalized. If confirmed, the order is logged into a Google Sheets orders sheet with a unique sequential number, and inventory is kept in sync. The system supports multiple languages, including Arabic, English, and French, and eliminates manual data entry.