This blog post discusses how Instacart transformed static grocery flyers into interactive shoppable experiences using advanced technologies like computer vision and large language models (LLMs). The post outlines their manual digitization process, challenges faced, and the development of a custom two-phase pipeline that significantly reduced processing time from hours to under thirty minutes per flyer. The approach leverages image segmentation techniques, OCR, and search algorithms to enhance the identification of products in digital flyers, illustrating a successful blend of various AI methodologies to improve retail experiences.