Skip to content
Ali Raza
All case studies

Multilingual AI Hotel Receptionist Voice Agent

A real-time voice receptionist that books rooms in Urdu, English, or Spanish and confirms them on WhatsApp.

Problem

A hotel has to answer booking calls in the caller's own language, then get each reservation recorded and confirmed.

What I built

  • Built a real-time voice receptionist on LiveKit Agents with Silero VAD, Groq Whisper-large-v3-turbo transcription, a Groq-hosted LLM, and ElevenLabs multilingual TTS, detecting the caller's language and replying in it (Urdu, English, Spanish).
  • Implemented LLM function calling for room booking that collects guest name, room type, and phone number, then posts to an n8n webhook with timeout and error handling so the agent recovers gracefully on failure.
  • Orchestrated n8n workflows that log every reservation to Google Sheets and send instant WhatsApp confirmations through a self-hosted Evolution API, plus a room-availability webhook returning status and pricing to the agent.

How a booking flows

  1. A caller speaks in Urdu, English, or SpanishLiveKit Agents + Silero VAD
  2. The agent detects the language and replies in itGroq Whisper + ElevenLabs
  3. It collects guest name, room type, and phone numberLLM function calling
  4. The reservation is loggedn8n + Google Sheets
  5. A WhatsApp confirmation goes out instantlyEvolution API

Result

A caller speaks Urdu, English, or Spanish, books a room by voice, and gets an instant WhatsApp confirmation. Every reservation is logged to Google Sheets, and the agent recovers gracefully if the booking webhook fails.

Need something like this built?

Book a 30-minute intro call and tell me about the role or the project.

Next case study: AI-Powered Workspace Management Platform