# Dmitry (Dima) Zabello

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Dmitry Zabello](https://cerebralvalley.ai/u/dzima)
- **GitHub:** https://github.com/foreverest/meta-hackathon-finger-pods
- **Demo video:** https://www.youtube.com/watch?v=U2252kDgJgw
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3978

Finger Pods explores a missing piece of the upcoming wearable computing era: a practical way to type.

Smart glasses, wearable AI, and increasingly compact hardware point toward a future where computing is available throughout the day without sitting behind a laptop or holding a phone. We already have cameras and microphones to capture the world, earbuds to deliver sound, and glasses that bring visual information into view. These technologies are coming together—but convenient text input remains a major gap.

Wearable computers need to support more than consuming information or issuing short voice commands. People also need to write messages, take notes, search, edit documents, and communicate with AI. If every substantial text interaction requires taking out a phone or opening a laptop, much of the promise of wearable computing is lost. We need text input that is portable, private, precise, and comfortable enough for everyday use.

Existing approaches involve significant compromises. Physical keyboards work well but are another device to carry and set up. Speech-to-text is useful, but speaking every message aloud is awkward in public, disruptive in quiet spaces, and unsuitable for private information. Camera-tracked virtual keyboards require the fingers to remain visible; systems using visible-light cameras also struggle in the dark. Tapping a bare surface provides no physical key feedback. Alternative wearable keyboards often require learning unfamiliar gestures or key combinations.

My approach is a keyboard you can wear, built around familiar typing movements and real tactile clicks. Finger Pods uses finger-worn clips containing physical buttons, magnets, and magnetic sensors. A trained machine-learning model combines the magnetic readings with the pressed button to recognize which key the finger is reaching for. The same button can produce different characters in different positions.

This approach works without cameras or dependence on lighting and aims to build on existing touch-typing muscle memory. The goal is to make entering text as natural a part of wearable computing as listening through earbuds.

The current one-hand prototype supports 17 keys and demonstrates live text entry. Next steps include improving comfort, testing across more users and sessions, supporting both hands, and going wireless.

## More from Meta Global AI Developer Hackathon

- [LineSight](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3975)
- [DocSyde](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3976)
- [MonkeeTools](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3977)
- [Prysm7](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3979)
- [Hardware Engineer](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3980)
- [VEYTAC](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3981)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/3978. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt