	In Browser
	StumbleUpon
	del.icio.us
	Google
	Google Buzz
	reddit
	LinkedIn

	Facebook
	Twitter
	Linkedin
	E-Mail

Generative AI > Google Gemini API > Real Estate Property Search Agent

Real Estate Property Search Agent

Author: Venkata Sudhakar

Property search in India involves juggling locality preferences, budget constraints, EMI calculations, and locality-specific market insights. A real estate agent can parse natural language requirements, search a property database, calculate home loan EMIs, and present a ranked shortlist with clear reasoning - exactly what a human broker does but instantly.

In this tutorial, we build a property search agent for the Mumbai real estate market. The agent understands buyer requirements, filters listings, calculates EMI, and generates a personalised comparison report.

The below example shows the agent handling a buyer query and producing a structured property shortlist.

import os
from google.adk.agents import LlmAgent
from google.adk.tools import FunctionTool
from google.adk.sessions import InMemorySessionService
from google.adk.runners import Runner
from google.genai import types

# Mock property database
PROPERTIES = [
    {"id": "P001", "name": "Lodha Palava", "location": "Dombivli", "city": "Mumbai",
     "bhk": 2, "area_sqft": 750, "price_lakhs": 65, "possession": "Ready",
     "amenities": ["gym", "pool", "clubhouse"], "metro_dist_km": 1.2},
    {"id": "P002", "name": "Hiranandani Estate", "location": "Thane", "city": "Mumbai",
     "bhk": 3, "area_sqft": 1100, "price_lakhs": 128, "possession": "Ready",
     "amenities": ["gym", "pool", "school", "mall"], "metro_dist_km": 0.5},
    {"id": "P003", "name": "Godrej Nest", "location": "Kandivali", "city": "Mumbai",
     "bhk": 2, "area_sqft": 680, "price_lakhs": 88, "possession": "Dec 2025",
     "amenities": ["gym", "garden"], "metro_dist_km": 0.3},
    {"id": "P004", "name": "Runwal Greens", "location": "Mulund", "city": "Mumbai",
     "bhk": 3, "area_sqft": 1050, "price_lakhs": 105, "possession": "Ready",
     "amenities": ["gym", "pool", "sports court"], "metro_dist_km": 0.8},
]

def search_properties(city: str, bhk: int, max_budget_lakhs: float, locality: str = "") -> dict:
    """Search property listings by city, BHK type, and budget."""
    results = [
        p for p in PROPERTIES
        if p["city"].lower() == city.lower()
        and p["bhk"] == bhk
        and p["price_lakhs"] <= max_budget_lakhs
        and (not locality or locality.lower() in p["location"].lower())
    ]
    return {"count": len(results), "properties": results}

def calculate_emi(principal_lakhs: float, rate_percent: float = 8.5, tenure_years: int = 20) -> dict:
    """Calculate home loan EMI. Assumes 80% LTV loan."""
    loan = principal_lakhs * 0.80 * 100000  # 80% of value
    r = rate_percent / 12 / 100
    n = tenure_years * 12
    emi = loan * r * (1 + r)**n / ((1 + r)**n - 1)
    return {
        "property_value_lakhs": principal_lakhs,
        "loan_amount_lakhs": round(principal_lakhs * 0.80, 2),
        "emi_per_month": round(emi),
        "rate_percent": rate_percent,
        "tenure_years": tenure_years
    }

property_agent = LlmAgent(
    name="real_estate_agent",
    model="gemini-2.0-flash",
    instruction="""You are a Mumbai real estate advisor.
Use search_properties to find matching listings and calculate_emi for each shortlisted property.
Present a ranked shortlist with: property name, location, price, EMI, pros, and cons.
Always mention metro distance and possession status.""",
    tools=[FunctionTool(search_properties), FunctionTool(calculate_emi)]
)

Now run the agent with a buyer query,

It gives the following output,

Property Shortlist - 2 BHK Mumbai under Rs 90 Lakhs

2 properties found matching your criteria:

--- RANK 1: Godrej Nest, Kandivali ---
Price: Rs 88 Lakhs | Area: 680 sq.ft
EMI: Rs 54,200/month (8.75%, 20 years, 80% LTV)
Possession: Dec 2025 | Metro: 0.3 km
Pros: Closest to metro, reputed developer, garden amenity
Cons: Under construction, smaller area, Dec 2025 possession risk

--- RANK 2: Lodha Palava, Dombivli ---
Price: Rs 65 Lakhs | Area: 750 sq.ft
EMI: Rs 40,100/month (8.75%, 20 years, 80% LTV)
Possession: Ready | Metro: 1.2 km
Pros: Lowest price, ready possession, gym/pool/clubhouse, larger area
Cons: 1.2 km from metro, Dombivli is further from central Mumbai

Recommendation: If metro proximity matters most, Godrej Nest is better.
If budget and immediate possession matter, Lodha Palava is the stronger choice.

Connect this agent to a real property database or MagicBricks/99acres API for live listings. Add a stamp duty and registration calculator tool for complete cost-of-purchase estimates. For a richer experience, add a neighbourhood scoring tool that rates areas by schools, hospitals, and transport connectivity.

Send your comments, suggestions or queries regarding this site to [email protected].