Abstract

A basic agent helping in Interior Design Consultation and Designing.

Introduction

The integration of intelligent agents into design applications has garnered significant interest in recent years, especially in fields like interior design. This project presents a novel approach using Agentic-AI, an AI-driven assistant designed to streamline consultation processes in interior design by leveraging the power of memory management, decision-making, and context-driven interactions. The application built in this project demonstrates how an AI agent can assist both professional designers and everyday users in creating well-informed design decisions through conversational interfaces. By integrating various AI libraries and tools, this prototype sets the foundation for more robust, context-aware AI agents
https://github.com/ethicalanp/Agentic-AI

Methodology

This study employs a structured experimental approach to evaluate the effectiveness of the Agentic-AI application. The agent leverages the Agno framework to provide memory-driven decision-making and context-aware suggestions. Key components of the system include:

Contextual Memory Management: The agent maintains a dynamic memory of interactions and designs, allowing it to make intelligent, context-based recommendations.

Data Integration: Data sources such as OpenAI and Groq APIs are integrated to enhance the agent's decision-making capabilities by providing embeddings and computational acceleration.

Consultation Framework: Through guided questioning and response evaluation, the agent supports both creative and engineering aspects of interior design, offering suggestions based on both aesthetic preferences and engineering constraints.
Libraries and frameworks used in this experiment include:

Agno: The core framework used for memory management and decision-making.

OpenAI: Integrated for generating embeddings and creating contextually relevant responses.

Groq: Used to speed up AI computations, leveraging specialized hardware for faster processing.

Other supporting technologies:

Python 3.12

Streamlit: For building interactive front-end interfaces.

Pandas: For data manipulation and analysis.

Python-dotenv: For managing environment variables.

PyPDF: For extracting data from design documents.

LanceDB: For managing large-scale data storage.

Tantivy: For full-text search indexing.

yFinance: To pull in financial data for budgeting and cost management.

Experiments

The experiment focused on two key aspects:

Design Consultation: The agent was tasked with generating design suggestions based on input from users about their space, preferences, and needs. The process involved guided questioning to narrow down preferences, such as color schemes, furniture styles, and layout options.

Engineering Feasibility: The agent also provided suggestions based on engineering requirements, considering aspects like room dimensions, structural limitations, and budget constraints. This phase also involved questioning to identify and address potential challenges in implementing the design.

The agent's decision-making process was evaluated in real-time by users interacting with the system, providing feedback on the suggestions provided. This feedback helped assess the relevance and practicality of the design recommendations.

Results

While the prototype is still in the early stages, the results from user interactions were promising. The AI was able to provide thoughtful, contextually relevant recommendations that balanced both creative design and practical constraints. The integration of Agno allowed the agent to adapt to each user's specific needs, while the OpenAI and Groq APIs provided fast, intelligent decision-making capabilities.

Despite the positive results, the system showed room for improvement in the areas of:

Contextual Awareness: The agent occasionally struggled to maintain long-term context across multiple design sessions.

Customization: While the agent could provide relevant suggestions, further personalization options could enhance its utility, especially for experienced designers.

Conclusion

Conclusion

This project demonstrates the potential of combining memory management and decision-making frameworks like Agno with advanced AI tools such as OpenAI and Groq to build an intelligent agent that aids in interior design. While the application is currently in its prototype phase and not a fully-fledged solution, it offers a strong foundation for future development. The integration of various libraries and APIs showcases how multiple data sources and computational tools can work together to enhance the agent's abilities, particularly in the context of design consultation and engineering feasibility.

Further development is required to refine the agent's memory management, contextual awareness, and customization features. However, the project has shown that intelligent agents can play a significant role in streamlining complex design processes, making them more accessible and efficient for both professionals and consumers alike.