In a single weekend, I built an AI-powered service that takes a screenshot of a food menu, extracts dish names using OCR (Tesseract), parses them with OpenAI, and generates realistic food images using a generative model. The project is open-source and live — check out the code and demo at the end of the article.