As of January 2025, India’s stock market had topped 110 million registered investors, indicating that retail participation is rapidly increasing. However, many retail investors lack a professional financial experience, making it difficult to understand sophisticated financial data and reports. This work presents a unified AI-powered platform with four fundamental functions to bridge this knowledge gap: (i) document-based question answering, (ii) peer comparison, (iii) automated report preparation, and (iv) portfolio optimization. The platform makes use of cutting-edge large language models, LLaMA 3.3 70B and LLaMA 4 Maverick 17B, which are integrated utilizing a Retrieval-Augmented Generation (RAG) architecture. These models evaluate structured Excel files and unstructured PDF financial reports to provide simple, actionable information. We use conventional measures like Cosine Similarity and ROUGE scores to evaluate model performance, and we find that LLaMA 4 Maverick 17B delivers the most correct answers, while LLaMA 3.3 70B balances precision and response consistency. The entire system is implemented using a FastAPI backend that generates professional, investor-friendly PDF summaries from raw financial data. Our findings support the viability and efficacy of generative AI in democratizing financial analysis, allowing individual investors with less financial competence to make informed investment decisions.