Driven by the rapid evolution of Generative AI, I was inspired to build Curator's Lens, a visual artificial intelligence tool that transforms historical archives into predictive narratives. I leveraged Google Gemini to generate the codebase and streamline the development process, building a Streamlit frontend powered by Groq’s LPU Inference Engine and the Llama 4 Scout (17B) vision model. This multi-model workflow allowed for sub-second identification of STEM historical figures while forecasting their technical legacy through 2050. Throughout the process, I went through many technical difficulties like shifting from various vision models and complex configurations, which helped me to refine my prompt engineering. This project allowed me to apply my prompt engineering skills to a spectacular technical masterpiece, proving that combining high-performance processing with professional design can turn a simple tool into a powerful interpretive experience.
Log in or sign up for Devpost to join the conversation.