Project 01 · Accessibility · AI / ML
Ishaara.
Making signs understood.
Sole frontend developer; backend and model collaboration
Translation interface, backend functionality, user analytics
Next.js / TensorFlow.js
01 / Context
The application.
Ishaara is an Indian Sign Language translator. The application uses camera-captured gestures and a trained recognition model to translate signs into text in real time.
02 / Implementation
What I built.
The application
As the sole frontend developer, I built the Next.js application and connected its interface to the translation workflow. I also developed backend functionality and an admin dashboard to track daily users.
The model integration
I collaborated with teammates to train a gesture-recognition model using camera-captured data, then integrated the model with the frontend through TensorFlow.js for real-time sign-to-text translation.
03 / Outcome
Reaching people through search.
The ishaara.vercel.app domain recorded 2.78K Google Search clicks and 37.7K impressions over three months, as reported in Google Search Console.
Google Search Console · ishaara.vercel.app · Jun. 2026–Sep. 2026 · Three-month reporting period in the resume.