<aside> 📌

This AI Agent handles all incoming messages from cold leads.

It answers their questions, qualifies them by asking smart follow-up questions, and guides them toward booking a demo or sales call.

In this example, it's set up for a real estate agency - but it can work for any business that deals with inbound leads.

With this system:

• Lead response time drops to seconds
• Qualification happens automatically
• No info gets lost — all data is saved to your CRM for follow-up
• Your sales team only talks to ready-to-buy leads

It’s like having a 24/7 SDR who never forgets anything.

</aside>

1. User sends a message

   Cold lead writes in the chat (website, WhatsApp, etc.).

2. Language is detected automatically

   The AI replies in the user’s native language.

3. Message is analyzed

   AI classifies what kind of question it is:

   • A general/process question? → goes to Knowledge Agent
   • A property or product request? → goes to Property Agent
   • A sales conversation? → goes to Sales Agent

4. Right AI takes over

   Depending on the message type:

   • Knowledge Agent answers questions about the company, legal processes, or buying steps (uses Qdrant vector search)
   • Property Agent shows real estate listings (pulled from Google Sheets)
   • Sales Agent runs a qualification script: asks for name, budget, goals, etc.

5. Data is saved in real time

   As soon as a lead shares their info, it’s saved to the database (Postgres).

6. AI remembers returning leads

   If a user comes back later, it picks up right where it left off - no repetition.

7. End goal

   The system pushes gently toward one thing: booking a call with a sales rep.

🧩 Tech Stack Under the Hood:

• n8n to connect all the parts
• OpenAI GPT-4o for smart understanding and replies
• Qdrant for document search (RAG)
• Google Sheets for live product data
• Postgres for lead tracking and memory