Inspiration
As passionate anime fans, we often browse sites like MyAnimeList and Crunchyroll, searching for the next series to binge. However, we often run into the issue of picking from a plethora of anime within a genre, with seemingly indistinguishable descriptions and synopses. Rather than tirelessly scouring for the perfect anime and poring over Reddit users' suggestions, we decided to leverage AI.
What it does
This webapp searches through anime media and filters them based on genre, user input, or type of media. The app also features an additional AI feature that can provide the closest related animes based on a user prompt.
How we built it
The webapp uses a database in postgreSQL from the website Anime World. The website was built using react and typescript for the frontend and express for the backend. For the AI search feature we utilized Gemini AI.
Challenges we ran into
Trying to craft the prompt such that the Gemini model we are using provides data in a usable format. Also attempting to retreive reliable results from the Gemini response and parsing them into usable JSON objects. We also struggled on hosting the postgreSQL database on AWS and connecting it to express.
Accomplishments that we're proud of
The AI functionality works and provides reliable and foused results based on the prompt provided by the user. Also, the app's UI is user-friendly and modern.
What we learned
How to use PostgreSQL databases, AWS cloud hosting, and effective file organization for React compoments.
What's next for AI-nime
User logins, such as bookmarking features. Also utilizing a more modern Gemini model so that the response time is shortened and the data retreived from Gemini is more focused.
Log in or sign up for Devpost to join the conversation.