GenAI for Investment Recommendations Using RSS Feed

This study presents the design and implementation of an advanced financial news analysis system aimed at providing real-time insights and responses to user queries. The system utilizes a multi-step pipeline that involves fetching live RSS (Really Simple Syndication) feed data from multiple sources, processing and categorizing the data, and applying natural language processing (NLP) techniques, such as named entity recognition (NER) and sentiment analysis, to extract and understand relevant information. To respond to user interactions, the system leverages LangChain and ChatGroq to generate contextually appropriate responses using a large language model (LLM). The resulting interface, built with Streamlit, facilitates intuitive user interaction and the seamless display of generated insights. Evaluation results demonstrate the system's high accuracy in categorizing news articles (average accuracy of 92%) and efficient response times for generating outputs (average response time of 4.75 seconds). The system's ability to handle incremental data updates, categorize news into relevant financial topics, and generate rapid, AI-driven responses makes it a powerful tool for financial analysis and decision-making.