Philer Inc. aims to enhance its Retrieval-Augmented Generation (RAG) chatbot to deliver personalized guidance for real estate transactions. While the current chatbot provides general information, this project focuses on refining its capabilities to offer relevant information based on user-specific scenarios.

This initiative will involve integrating data sources to improve accuracy, enhancing natural language processing (NLP) to handle queries, and expanding coverage to include various real estate transaction types, such as buying, selling, and refinancing. Additionally, the chatbot will be designed to ensure compliance with legal standards, making it a reliable and valuable tool for users navigating real estate transactions.

The project will analyze existing chatbot functionalities, identify gaps, and implement improvements to enhance response accuracy and relevance. Testing will be conducted to ensure the chatbot performs effectively in real-world scenarios.

By the end of the project, the chatbot will be capable of delivering precise, context-aware transaction support in real estate closings. This advancement will elevate user experience and improve operational efficiencies, reducing the burden on customer support.

Technology Stack

• Hosting: AWS
• Frontend: React.js (Tailwind CSS)
• Backend: Python (FastAPI, LangChain)
• Vector Database: Pinecone
• LLM & Embeddings: OpenAI (GPT-4-turbo)