Finding a book you want to read isn’t always a quick Google search away. Sometimes a conversation with a friend can suss out the important nuances of the stories and ideas you’re interested in. But not all of your friends have a vast body of knowledge of all the books that are out there. This web app project explored the possibilities of a chatbot that makes books recommendations after having a conversation with you.

After your converstation, books are recommended based on the conversation content and results from the Google Books API.

User tests with friends revealed some differing opinions on chatbots. Some doubted if they needed an AI for book recommendations. One friend remarked, “This may not be the most efficient way to find great books, but it can be a fun way to find books I like.”

Responsive design for mobile interfaces were developed.

Early iterations focused on the responsive design layout and typography, while later iterations added a warm color palette to counteract any sci-fi hi-tech connotations of the experience.

Last but not least, the logo was also designed to resemble other chat interfaces and reflect the project/chatbot’s name.

Side Note: At the time we did this project, Omni only asked a preset list of questions. Large language models were commonly found in research and some early commercial APIs, though not at the performance and usability of ChatGPT. A part of me wonders how this would have turned out with a generative AI asking questions rather than a preset list of questions…

Technical Skills

Illustrator, Invision, ReactJS, Webpack

Team

Frontend/UX (ReactJS): Dan Lu
Backend (PostgreSQL/AWS/ReactJS): Robert Chen, Wences Lee