🏡 Propnesto Real Estate AI Voice Consultant ("Ananya")

Production-Grade Autonomous AI Calling System & Telephony CRM for Real Estate Lead Qualification & Site Visit Automation (Dograh Agent Builder Configuration).

---

🎙️ Dograh Agent Builder System Prompt (Copy & Paste)

You are Ananya, an expert Property Consultant from Propnesto Real Estate.
STRICT CONSTRAINT: Keep the entire call duration under 60 seconds (1 minute max).

CONVERSATION RULES:
1. Speak naturally, warmly, and at a crisp pace in English or Hindi based on the customer.
2. Keep EVERY utterance under 15 words. Ask ONE question at a time.
3. Greeting: "Hello! This is Ananya from Propnesto Real Estate. Am I speaking with {{name}}? Got 60 seconds regarding to let you know about your nearby property projects?"
4. If Busy: "No problem! What date and time can I call you back?" -> Save {{callback_time}} and end.
5. If Free:
   - Ask Intent: "Are you looking to Buy, Rent, Sell, or Invest?" -> Extract {{intent}}
   - Ask Purpose: "Is this for Personal Use or Investment?" -> Extract {{purpose}}
   - Qualify Rapidly: City {{city}}, BHK {{bhk}}, Budget {{budget}}, Possession {{possession}}.
   - Match: "We have matching {{bhk}} options in {{city}}. Shall I book a Morning site visit for you?"
   - Confirm Details: Name {{name}} & WhatsApp number {{whatsapp_number}}.
6. Score Lead:
   - HOT: Site visit booked or buying immediately. -> {{lead_score}} = "Hot"
   - WARM: Interested within 1-3 months. -> {{lead_score}} = "Warm"
   - COLD: Casual inquiry. -> {{lead_score}} = "Cold"
7. Wrap Up within 60 Seconds: "Thank you {{name}}! Sending property brochures on WhatsApp right now. Have a great day!"

---

🔑 Dograh Dynamic Variable Extraction Schema

Paste these variable names into your Dograh Variable Settings & Dynamic Extraction:

Variable Name	Type	Description	Extraction Trigger
name	String	Customer Full Name	Extracted from greeting confirmation / user response
city	String	Target City	Extracted from qualification ("Gurgaon", "Noida", "Mumbai")
intent	String	Real Estate Intent	Buy | Rent | Sell | Investment
purpose	String	Property Usage Purpose	Personal | Investment
property_type	String	Property Category	Apartment | Villa | Plot | Commercial
bhk	String	Bedroom Configuration	1 BHK | 2 BHK | 3 BHK | 4 BHK | N/A
budget	String	Budget Bracket	Under ₹25L | ₹25–50L | ₹50L–1Cr | Above ₹1Cr
possession	String	Possession Requirement	Ready to Move | Under Construction
timeline	String	Purchase Timeline	Immediately | Within 1 month | 3 months
loan	String	Loan Required	Yes | No
whatsapp_number	String	WhatsApp Contact Number	Extracted from detail confirmation
site_visit_requested	String	Site Visit Status	Yes | No
visit_slot	String	Visit Time Slot	Morning | Evening
callback_time	String	Requested Callback Time	Extracted if customer is busy
lead_score	String	Automated CRM Score	Hot | Warm | Cold

---

🤖 Dograh Post-Call Extraction JSON Schema (Copy & Paste)

Paste this JSON schema into Dograh's Post-Call Data Extraction Hook:

{
  "type": "object",
  "properties": {
    "name": { "type": "string", "description": "Customer full name" },
    "city": { "type": "string", "description": "Target city" },
    "intent": { "type": "string", "enum": ["Buy", "Rent", "Sell", "Investment"] },
    "purpose": { "type": "string", "enum": ["Personal", "Investment"] },
    "property_type": { "type": "string", "enum": ["Apartment", "Villa", "Plot", "Commercial"] },
    "bhk": { "type": "string", "description": "1 BHK, 2 BHK, 3 BHK, 4 BHK" },
    "budget": { "type": "string", "description": "Customer budget" },
    "possession": { "type": "string", "enum": ["Ready to Move", "Under Construction"] },
    "whatsapp_number": { "type": "string", "description": "WhatsApp phone number" },
    "site_visit_requested": { "type": "string", "enum": ["Yes", "No"] },
    "visit_slot": { "type": "string", "enum": ["Morning", "Evening"] },
    "callback_time": { "type": "string", "description": "Convenient callback time if busy" },
    "lead_score": { "type": "string", "enum": ["Hot", "Warm", "Cold"] }
  },
  "required": ["intent", "lead_score"]
}

---

🚀 Running the System

# 1. Start Backend Server (Port 5001)
cd backend && npm start

# 2. Start Frontend App (Port 3000)
cd frontend && npm run preview -- --port 3000

Open http://localhost:3000 in your browser.