Shelf Help is a book recommendation prototype built during Viget's internal hackathon to address the shortcomings of tools like Goodreads and Amazon. It uses a plain-language AI interface to let readers describe exactly what they want, including genre, sub-genre, mood, prose style, and length, while also importing a Goodreads library CSV to exclude already-read books. The app was designed in Figma and converted to a functional prototype using Base44, a no-code AI app builder. Key features include tailored recommendation results, a 'mark as read' option, thumbs up/down feedback, and a search history called 'The Stacks.' Limitations encountered include hallucinated book titles, outdated recommendations, and inconsistent prompt adherence from the AI builder.
Sort: