TWINKLE

Inspiration

We noticed that most productivity tools are built to optimize output — but none of them ask how you're actually feeling. Watching Gen Z struggle silently with burnout, loneliness, and emotional overwhelm while trying to study or work, we asked: what if a tool could sense your emotional state before you even realize it yourself? That question became Twinkle.

What it does

Twinkle is a voice-powered emotional readiness tool for students. It checks in with you through a voice AI avatar, detects your emotional state from tone and behavior, and adapts its support in real time, offering encouragement, pacing adjustments, or lighter tasks. It turns your emotional readiness into a measurable score, so you can study smarter without ignoring how you feel.

How we built it

We built Twinkle using voice emotion detection models that analyze pitch, tone, and frequency to assess emotional state in real time. The app combines a conversational AI layer for check-ins, an adaptive recommendation engine for study support, and a personal dashboard to track emotional patterns over time.

Challenges we ran into

Detecting nuanced emotional states from voice alone, – beyond simple happy/sad labels, – proved far more complex than expected. Balancing accuracy with a non-intrusive, user-friendly experience was a constant tension, as was designing a system that feels supportive rather than surveillance-like.

Accomplishments that we're proud of

Building a real-time voice emotional readiness score that actually responds meaningfully to different emotional states. Creating a UX flow that feels warm and human. And proving that emotional intelligence and productivity tools don't have to be separate categories.

What we learned

Voice is one of the most honest signals we have, people can mask their words, but rarely their tone. We also learned that emotional support in a tool needs to feel like a conversation, not a diagnostic. Trust and safety in the UX matters as much as the underlying technology.

What's next for TWINKLE

Deeper personalization through longitudinal emotional tracking, integration with Pomodoro and focus apps, expansion beyond students into workplace wellness, and, eventually, building a voice model trained specifically on Gen Z emotional patterns to make Twinkle's detection more accurate and culturally resonant.

Built With

• figma
• vector