Academic advising is essential for student success, encompassing a wide range of inquiries, from general program information to more personalized concerns. To streamline the process and limit human interaction in this work, we introduce an innovative AI-driven chatbot that harnesses the power of large language models (LLM), specifically LLaMA, along with retrieval-augmented generation (RAG) and fine-tuning (FT) techniques. This approach aims to automate and improve the student advising process, offering more efficient and accurate support for students. Unlike conventional chatbots constrained by rigid, intent-based frameworks, our system dynamically retrieves and synthesizes information from continuously updated knowledge bases. This approach enables the chatbot to provide accurate, context-aware, and reliable responses to student inquiries. Compared to existing systems, the proposed solution effectively mitigates common issues such as hallucinations and outdated information, providing factual, grounded, and scalable performance. This paper presents an overview of the development lifecycle, encompassing dataset creation, system architecture, and experimental validation. Our experiments show that Retrieval-Augmented Generation (RAG) outperforms Fine-Tuning in accuracy, scalability, and relevance for academic advising chatbots. RAG improves the handling of individual queries, automates routine tasks, and reduces the advisor’s workload, allowing them to focus on complex cases. This solution improves student support by balancing efficient automation with personalized guidance.