2025-06-032025-06-032025-05-19https://repository.tcu.edu/handle/116099117/67166Academic advising is a critical yet time-intensive process that often leaves students with limited access to timely, personalized guidance. This thesis introduces Echelon, an AI-powered academic advisor designed to augment the advising experience by leveraging semantic embeddings, large language models (LLMs), and retrieval-augmented generation (RAG). Echelon ingests institutional documents and student transcripts to deliver accurate, context-aware responses through a conversational web interface. Echelon is publicly accessible via echelongpt.com and can also be internally deployed on TCU's servers, offering enhanced privacy and security controls tailored for institutional use. This paper details Echelon's underlying technologies, system architecture, and deployment and outlines future enhancements such as proactive advising and domain expansion beyond academia. The result is a scalable, extensible framework for transforming how students interact with institutional knowledge.AIMLLLMChatGPTAdvisingRustPython