Handshape is designed to be the definitive digital resource for American Sign Language. Rather than treating ASL as a secondary translation layer mapped to English, Handshape builds its entire interaction model around visual discovery and spatial learning.

ASL-native phonology model — handshape · location · movement — powering search
Real-time hand tracking via Apple's Vision framework: 21 landmarks, on-device, nothing recorded
Serverless AWS backend (Cognito, DynamoDB, MediaConvert) — all Terraform

I built Handshape to empower both native signers and learners, fostering a deeper appreciation for the language's rich visual grammar.

Overview Architecture