The past two weeks in ISM have been a turning point for my project. In addition to making technical progress, I had the opportunity to interview Mr. Veera Vemula, an industry professional who has agreed to mentor me throughout this journey. Our conversation was eye-opening, as he provided invaluable feedback on my AI-driven skin cancer detection app and introduced me to advanced techniques that could significantly enhance its accuracy and functionality. One of the key takeaways from our discussion was the potential of incorporating Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs) into my project. RAG is an advanced AI technique that enables models to retrieve relevant information from external sources in real time, enhancing their decision-making capabilities. This could be a game-changer for my app, allowing it to pull in relevant medical data and research while analyzing skin lesion images. Mr. Vemula’s suggestion to integrate LLMs into my app opened up exciting new possibilities. By adding an interactive dimension, the app could provide a diagnosis and explain its reasoning to users in an understandable way. This means that instead of just receiving a probability score, users could get detailed insights into why a particular skin abnormality may be concerning, along with suggestions for next steps. This shift transforms my app from being a simple detection tool to a more comprehensive health companion.
