Software · CS Capstone · Asian Art Museum

Musea

personalized AI museum tours, built with the Asian Art Museum

For my Computer Science capstone, I worked with three teammates in collaboration with the Asian Art Museum in San Francisco to make the museum experience more intentional, personal, and creative for young people. We created Musea, an open-source personalized tour web app that generates custom museum routes and artwork recommendations based on a visitor’s searched words, interests, and previously liked artworks.

Musea brings together six core features into one self-directed visit: semantic search, path generation, an audio guide, personalized recommendations, collections, and a “talk to me” conversation mode.

One of the highlights was traveling to Amsterdam with my team to demo Musea to the Stedelijk Museum and the Mauritshuis. The conversations and feedback we received helped us refine our core value proposition: personalized tour generation directly on museum maps, allowing visitors to move through a collection in a way that feels self-directed.

I worked across the stack on audio guide generation, semantic search, personalized recommendations, the collections feature, and pathfinding. Through the project, I learned a lot about both backend and frontend development, including building recommendation pipelines, working with artwork metadata and embeddings, and turning ambiguous user feedback into concrete product decisions.

More broadly, Musea taught me how to balance technology supporting cultural exploration without replacing the wonder of being in a museum.