AI WhatsApp Lead Qualifier for Real Estate with Gemini, Airtable & Calendar

Qualify Real Estate Leads on WhatsApp—Automatically with Gemini, Airtable & Google Calendar

Turn incoming WhatsApp messages into actionable real estate leads. This n8n workflow uses an AI WhatsApp assistant to answer property questions from a live Airtable database, classify leads as hot, warm, or cold, and check Google Calendar availability before helping book viewings—escalating to a human agent only when it’s time.

What this workflow does

• Receives WhatsApp messages via a WhatsApp Trigger when customers send a text or voice note.
• Routes and processes input: text messages are handled directly; voice notes are downloaded and transcribed using Google Gemini first.
• Answers from your live listings by querying Airtable (your selected Properties base and table) to look up real property data.
• Checks scheduling availability using Google Calendar before proposing or preparing viewing times.
• Returns structured results with a reply plus lead status (hot/warm/cold) and whether escalation to a human is needed.
• Sends the reply back to the customer and logs the lead to Google Sheets.
• Notifies your agent when ready for hot leads (duplicate alerts prevented). If the AI fails, the customer still receives a fallback reply while the developer is alerted separately.

Use cases

• A prospect asks about availability, pricing, or features—your AI answers instantly from Airtable.
• A voice note comes in with viewing intent—Gemini transcribes, qualifies the lead, then checks Calendar availability.
• Hot leads trigger a human notification, while warm/cold inquiries continue automated follow-up logic.

Technical details

• WhatsApp Business Cloud API: WhatsApp Trigger, Send WhatsApp Reply, Notify Agent, and Send Unsupported Notice.
• Google Gemini: Gemini Chat Model and Analyze Audio for voice transcription.
• Airtable: search_properties tool using your Properties base/table.
• Google Calendar: availability checks for viewings.
• Workflow logic uses nodes such as if, set, code, switch, and sticky note, plus structured output parsing.