MultiModal Agentic Framework to Create Real Estate Brochures

Multimodal agentic frameworks represent a cutting-edge approach in artificial intelligence, integrating various data types—such as text, images, audio, and video—to enhance the capabilities of intelligent systems. These frameworks utilize intelligent agents that can autonomously process and analyze diverse information sources, enabling more nuanced understanding and decision-making. By combining multimodality with agentic functionalities, these systems can adapt in real time to dynamic environments and user interactions. This integration not only improves operational efficiency across industries but also enriches human-computer interactions, making them more intuitive and context-aware. As such, multimodal agentic frameworks are poised to transform how we engage with technology in numerous applications.

Learning Objectives

• Understanding Agentic AI with Image Generation
• Exploring Camel AI Functionalities
• Developing a Multimodal Agentic System with CAMEL AI
• Benefits to Real Estate Businesses

This article was published as a part of the Data Science Blogathon.

MultiModal Agentic AI: Agents with Image Generation

Agentic AI represents a significant evolution in artificial intelligence, characterized by its autonomy and advanced decision-making capabilities. Integrating Agentic Frameworks with Image Generation capabilities can give significant advantages as mentioned below –

• Enhanced Creativity: These systems can assist in creative processes by generating unique visual content, enabling artists, designers, and marketers to explore new ideas and concepts efficiently.
• Personalization: By generating tailored images based on user preferences or data inputs, agentic systems can create personalized experiences in marketing, advertising, and entertainment.
• Rapid Prototyping: Agentic systems can quickly produce visual prototypes for products or concepts, facilitating faster iterations and feedback during the design process.
• Data Visualization: They can transform complex data sets into intuitive visual representations, aiding in better understanding and communication of information across various fields such as business analytics and scientific research.
• Accessibility: These systems can democratize access to high-quality visual content, allowing individuals and organizations without extensive design resources to create professional-grade images.
• Automation of Repetitive Tasks: By automating the image generation process, agentic systems reduce the time and resources spent on routine design tasks, allowing human creators to focus on more strategic initiatives.

What is Camel AI?

Camel AI (short for Communicative Agents for Mind Exploration of Large-Scale Language Model Society) is an innovative framework dedicated to the development and research of autonomous, communicative agents. Its primary goal is to examine how AI systems interact and collaborate, reducing the need for human involvement in various tasks. Focusing on the analysis of behaviors, abilities, and potential risks within multi-agent systems, Camel AI is an open-source project designed to foster collaboration and drive innovation within the AI research community.

Core Modules in Camel AI

The CAMEL framework is designed for the creation and management of multi-agent systems, incorporating several key components. It includes Models for defining agent intelligence, Messages for communication, and Memory systems for data storage and retrieval. The framework also integrates Tools for specialized tasks, Prompts to guide agent behavior, and Tasks to manage workflows. The Workforce module enables the formation of agent teams for collaboration, while the Society module facilitates interaction among agents. Together, these components enable the development of dynamic, collaborative multi-agent environments.

Toolkits in Camel AI

One of the greatest pros of using Camel AI is its integration with a diverse set of toolkits which can be seamlessly leveraged in creating multi-agentic systems. Camel AI includes several toolkits that enhance the capabilities of its multi-agent framework. Key toolkits include:

• Function Tool: This toolkit allows agents to call functions and interact with various APIs, facilitating complex task execution and integration with external services.
• Reddit Toolkit: This toolkit enables agents to interact with the Reddit API, allowing them to collect top posts, perform sentiment analysis on comments, and monitor discussions across subreddits.
• Retrieval Toolkit: Designed for information retrieval, this toolkit allows agents to query local vector storage systems, retrieving relevant information based on user queries.
• Media Tools: This includes functionalities for processing images and audio, enabling agents to handle multimedia content effectively.
• Document Tools: This toolkit provides capabilities for processing documents in various formats (e.g., PDF, Word) and includes web scraping features.
• Web Tools: These tools enable agents to access and interact with web services, such as search engines and APIs like DuckDuckGo and Wikipedia.
• DALL-E Integration: Camel AI also supports integration with image generation models like DALL-E, allowing agents to create images based on textual descriptions, enhancing their creative capabilities.
• Search Toolkits. A toolkit for performing web searches using various search engines like Google, DuckDuckGo, Wikipedia, and Wolfram Alpha.

These toolkits collectively empower Camel AI to perform a wide range of tasks, from data retrieval and processing to multimedia handling and creative image generation.

DALL-E

DALL-E is a series of advanced text-to-image models developed by OpenAI that generate digital images based on natural language descriptions, known as prompts. The initial version was released in January 2021, followed by DALL-E 2 in 2022, and the latest iteration, DALL-E 3, was integrated into ChatGPT and made available in late 2023.

DALL-E can create images in various styles, including photorealistic images and artistic renditions. It can manipulate and rearrange objects within images and infer details not explicitly mentioned in prompts.

Hands-On Implementation of a Multi-Modal Agentic System

In the following hands-on tutorial, we create a multi-modal agentic system using CAMEL AI for designing brochures for upcoming real estate projects in a city. This could help real estate businesses immensely as this aids in the automated creation of the brochures needed for giving out to clients when any of their new projects come up in a city without minimal human intervention.

Step 1. Installation of Necessary Libraries

!pip install 'camel-ai[all]'

Step 2. Defining Open AI API Keys

import os
os.environ['OPENAI_API_KEY'] = ''

Step 3. Importing Necessary Libraries

from camel.agents.chat_agent import ChatAgent
from camel.messages.base import BaseMessage
from camel.models import ModelFactory
from camel.societies.workforce import Workforce
from camel.tasks.task import Task
from camel.toolkits import (
    FunctionTool,
    GoogleMapsToolkit,
    SearchToolkit,
)
from camel.toolkits import DalleToolkit

from camel.types import ModelPlatformType, ModelType

import nest_asyncio
nest_asyncio.apply()

Step 4. Defining the Agents

search_toolkit = SearchToolkit()
search_tools = [
        FunctionTool(search_toolkit.search_duckduckgo)]
    


#Define the Model for the Agent as well. Default model is "gpt-4o-mini" and model platform type is OpenAI
guide_agent_model = ModelFactory.create(
        model_platform=ModelPlatformType.DEFAULT,
        model_type=ModelType.DEFAULT,
    )  

#Defining the Real Estate Agent for crafting the brochures
real_estate_agent = ChatAgent(
        BaseMessage.make_assistant_message(
            role_name="Real Estate Specialist",
            content="You are a Real Estate Specialist who is an expert in creating Description of Upcoming Residential Projects",
        ),
        model=guide_agent_model,
    )

#Defining the Agent for Real Estate Property Names
property_title_agent = ChatAgent(
        BaseMessage.make_assistant_message(
            role_name="Real Estate Project Name Specialist",
            content="You are a Real Estate Project Name Specialist who is an expert in Generating Trendy Names FoR Residental Projects in india",
        ),
        model=guide_agent_model,
    )


#Defining the agent for generating all the amenities near a location
location_benefits_agent = ChatAgent(
        BaseMessage.make_assistant_message(
            role_name="Real Estate Location Specialist",
            content="You are a Real Estate Location Specialist who is an expert in Generating All the amenities like malls, airports, markets, metro stations, railway stations etc with distances from a location of the mentioned property",
        ),
        model=guide_agent_model, tools =search_tools
    )


#Define the web search tool for the Agent using Tavily (we need to define the Tavily API Key beforehand)
dalletool = DalleToolkit()
imagegen_tools = [
    FunctionTool(dalletool.get_dalle_img),
    
]

#Define the Image Generation Agent with the pre-defined model and tools and Prompt
image_generation_agent = ChatAgent(
        system_message=BaseMessage.make_assistant_message(
            role_name="Image Generation Specialist",
            content="You can Generate Images For Upcoming Real Estate Projects For Showing to Clients",
        ),
        model=guide_agent_model,
        tools=imagegen_tools,
    )

This code snippet defines several agents using a model factory and a chat agent framework.

1. Model Creation: It first creates a default model (guide_agent_model) for the agents, specifically using the “GPT-4o-mini” model from OpenAI.
2. Real Estate Agents: Two agents are instantiated: one as a “Real Estate Specialist” focused on creating descriptions for upcoming residential projects, and another as a “Real Estate Project Name Specialist” tasked with generating trendy names for residential projects in India.
3. Real Estate Location Specialist : This agent is for generating all the amenities like malls, airports, markets, metro stations, railway stations etc with distances from a location of the mentioned property
4. Image Generation Tool: An image generation tool (dalletool) which allows the agents to generate images related to real estate projects.
5. Image Generation Agent: Finally, an “Image Generation Specialist” agent is created, equipped with the previously defined model and image generation tools to create visuals for upcoming real estate projects to present to clients.

Step 5. Defining the WorkForce

#Define the workforce that can take case of multiple agents
workforce = Workforce('Real Estate Brochure Generator')
workforce.add_single_agent_worker(
        "Real Estate Specialist",
        worker=real_estate_agent).add_single_agent_worker(
        "Real Estate Project Name Specialist",
        worker=property_title_agent).add_single_agent_worker(
        "Location Amenity Specialist",worker=location_benefits_agent).add_single_agent_worker(
        "Image Generation Specialist",
        worker=image_generation_agent)

 # specify the task to be solved Defining the exact task needed
human_task = Task(
        content=(
            """Craft a Brochure Content For a Upcoming Residential Real Estate Project in Sector 47,Gurgaon. The content should contain all the types of flats it has, all amenities in it and other such necessary details . 
            Provide a Name for this Property as well.
            Generate all the amenities of the location (with respect to its proximity to all public places) to this brochure content.
            Generate an Image of this Upcoming Project as well."""
        ),
        id='0',
    )
task = workforce.process_task(human_task)

This code defines a “workforce” that manages multiple agents for generating a real estate brochure. It adds 4 agents: a Real Estate Specialist, a Property Name Specialist, Location Amenity Specialist and an Image Generation Specialist. Then, it specifies a task for the workforce to complete—creating brochure content, providing a project name, and generating an image for a new real estate project in Gurgaon. The workforce processes the task by coordinating the agents to execute their respective roles.

Outputs

1. Output from Brochure Content Agent

Upcoming Residential Project in Sector 47, Gurgaon

Welcome to Your New Home
Discover the perfect blend of luxury and comfort in our upcoming residential
 project located in the heart of Sector 47, Gurgaon. Designed to cater to
 diverse lifestyles, our project offers a variety of flats that promise to
 meet your needs and exceed your expectations.

---

Flat Types Available:

1. **1 BHK Flats**
- **Size:** 600 sq. ft.
- **Description:** Ideal for young professionals or couples, these cozy 1 BHK
 flats feature an open living area, a modern kitchen, and a comfortable
 bedroom. Enjoy a well-designed space that maximizes functionality without
 compromising on style.

2. **2 BHK Flats**
- **Size:** 1,200 sq. ft.
- **Description:** Perfect for small families, our 2 BHK flats offer spacious
 living areas, two well-appointed bedrooms, and ample storage. Experience a
 harmonious blend of elegance and practicality, with large windows that
 invite natural light into your home.

3. **3 BHK Flats**
- **Size:** 1,800 sq. ft.
- **Description:** Designed for larger families, these expansive 3 BHK flats
 provide generous living spaces, three bedrooms, and a modern kitchen. Enjoy
 the luxury of space and comfort, with thoughtfully designed layouts that
 cater to your family’s needs.

4. **Penthouse Suites**
- **Size:** 2,500 sq. ft.
- **Description:** Elevate your living experience with our exclusive
 penthouse suites. Featuring stunning views, expansive terraces, and high-end
 finishes, these luxurious homes are perfect for those who appreciate the
 finer things in life. Enjoy private outdoor spaces and a lifestyle of
 sophistication.

---

 Amenities:
- **Clubhouse:** A state-of-the-art clubhouse with recreational facilities.
- **Swimming Pool:** Relax and unwind in our beautifully designed pool.
- **Gymnasium:** Stay fit with our fully equipped gym.
- **Landscaped Gardens:** Enjoy serene green spaces for relaxation and
 leisure.
- **24/7 Security:** Ensuring your safety and peace of mind.

---

Location Benefits:
- Proximity to major schools, hospitals, and shopping centers.
- Excellent connectivity to Delhi and other parts of Gurgaon.
- A vibrant neighborhood with parks, restaurants, and entertainment options.

---

Conclusion:
Don’t miss the opportunity to be a part of this exceptional residential
 community in Sector 47, Gurgaon. Whether you are looking for a cozy 1 BHK or
 a luxurious penthouse, we have the perfect home waiting for you. For more
 information and to schedule a visit, contact us today!

2. Output from Real Estate Project Name Specialist Agent

**Gurgaon Heights**

Output from Location Amenity Specialist Agent

Amenities and Proximity to Public Places near Gurgaon Heights, Sector 47,
 Gurgaon

1. Shopping Malls:-

**Ambience Mall** - 5 km
- **DLF Mega Mall** - 4.5 km
- **Sahara Mall** - 6 km

2. **Metro Stations:**
- **Huda City Centre Metro Station** - 4 km
- **Sikandarpur Metro Station** - 7 km

3. **Railway Stations:**
- **Gurgaon Railway Station** - 8 km
- **New Delhi Railway Station** - 30 km

4. Airports: -

**Indira Gandhi International Airport** - 15 km

5. Schools: -

**The Shri Ram School** - 2 km
- **G.D. Goenka Public School** - 3 km
- **Delhi Public School, Sector 45** - 3.5 km

6. Hospitals:-

**Medanta - The Medicity** - 6 km
- **Fortis Memorial Research Institute** - 5 km
- **Max Hospital, Gurgaon** - 7 km

7. Parks and Recreation:-

**Aravali Golf Course** - 3 km
- **Leisure Valley Park** - 4 km
- **Sukhna Lake Park** - 5 km

8. Restaurants and Cafes:-

**Cyber Hub** - 6 km
- **Sector 29 Food Street** - 5 km
- **The Great India Place** - 7 km

9. Entertainment: -

**PVR Cinemas, Ambience Mall** - 5 km
- **Kingdom of Dreams** - 8 km

3. Output From Image Generation Specialist:-

Conclusion

In conclusion, the integration of agentic AI systems with image generation capabilities, such as those found in the Camel AI framework (MultiModal Agentic Framework), represents a transformative advancement in both creativity and automation. By combining the power of autonomous decision-making with advanced image generation tools, these systems offer significant potential for rapid prototyping, personalized experiences, and enhanced accessibility to high-quality visual content. As Camel AI (MultiModal Agentic Framework) continues to evolve, it can drive innovation across various industries, reducing human involvement in routine tasks while empowering more strategic and creative endeavours.

Key Takeaways

1. Autonomous Creativity: Agentic AI systems with image generation capabilities enhance creative processes, allowing artists and designers to quickly generate unique and innovative visual content.
2. Personalized Experiences: These systems can tailor images based on user preferences, enabling customized marketing, advertising, and entertainment experiences.
3. Efficient Prototyping: Agentic AI accelerates the prototyping process by generating visual prototypes rapidly, fostering quicker iterations and feedback in design workflows.
4. Data Visualization: Agentic AI systems can convert complex data into clear, visually intuitive representations, aiding in better understanding and communication across diverse fields.
5. Multi-Agent Collaboration: Camel AI’s framework promotes collaboration among autonomous agents, enhancing task execution and facilitating the development of advanced, multi-agent systems for a wide range of applications.

The media shown in this article is not owned by Analytics Vidhya and is used at the Author’s discretion.

Frequently Asked Questions

Nibedita Dutta

Nibedita completed her master’s in Chemical Engineering from IIT Kharagpur in 2014 and is currently working as a Senior Data Scientist. In her current capacity, she works on building intelligent ML-based solutions to improve business processes.