Speak with an Elephant

I used a Python script to cut up the audio files into the 212 distinct sounds using the spreadsheet in an .csv format. I used another Python script to filter out certain frequencies that would be out of range for an elephant. The script also filtered out background static and adjusted the volumes. I wrote another Python script to group distinct types of sounds into clusters. I assigned meanings to each of the words using this organization. I used Claude API to translate English to elephant using the dictionary of sounds I created. I made a React.js application as a way to present the app and a Node.js application as a way to call the API. I used Affinity Designer to create the simple vector drawings. I used Google Cloud Run to host the Node project and Netlify to host the website, connecting it to a subdomain of my personal website.

Built With

• claude
• node.js
• python
• react.js
• typescript