← Back to work
AI · Accessibility · Hackathon · 2026

Sign2Graph

An AI-powered STEM learning platform that combines webcam-based air writing, gesture recognition, mathematical-input classification, animated explanations, and accessibility-oriented learning features.

Operating context

Sign2Graph won 3rd place overall at STEMist Hacks IV, an international hackathon with 400+ registered participants. The event offered $2,500 in cash and promoted $30,000+ in credits and subscriptions.

Problem

Mathematical input and visual explanation can be barriers for learners who do not thrive in conventional keyboard-first interfaces.

Approach

Combined webcam-based air writing, gesture recognition, input classification, animated explanations, and accessibility-oriented learning features into one STEM learning workflow.

Key product decisions

  • Made air writing and gesture recognition part of the core input path, not a cosmetic addition.
  • Paired recognized input with animated explanation to make the next step clearer.
  • Used accessibility-oriented interaction as a product requirement from the beginning.

Technical architecture

The platform joins live camera input with gesture and mathematical-input interpretation before presenting a learning-focused visual response.

Selected numbers

  • 3rd place overall
  • 400+ registered participants
  • $2,500 cash offered
  • $30,000+ credits and subscriptions promoted

Current result

3rd place overall at STEMist Hacks IV. T-Mobile was the presenting sponsor; the judging and workshop panel included professionals from Microsoft, Walmart Global Tech, Disney, Oracle, Electronic Arts, and Caterpillar.