🌐 Live Demo: https://ecarebots.com
🏁 Submission: No Keyboards Allowed Hackathon

EcareBots is a hands-free, multi-modal care platform for users who struggle with keyboards, mice, or touchscreens.
Interact using voice 🗣️ and gestures ✋ for vital actions: alerts, reminders, queries — perfect for elderly, disabled, or anyone in assisted-living or tough accessibility scenarios.

🏥 Traditional input = friction. For those with mobility, vision, or motor challenges, ordinary computers lead to dependence and frustration.
EcareBots restores independence and control, so users perform daily tasks simply by speaking or moving — and thrive with new freedom.

ecarebots/ ├── README.md # 📘 Project documentation ├── package.json # ⚙️ Dependencies & scripts ├── index.html # 🏠 App shell ├── vite.config.js # 🚀 Build setup └── src/ ├── App.jsx # 🎨 Main UI & logic ├── main.jsx # 🏁 Entry point └── styles.css # 🎀 Styles

text

+-------------------------+ | 👀 Frontend (Browser) | +-------------------------+ | Mic 🎤 / Webcam 📹 v +------------------------------+ | 🧠 Speech-to-Text (Web API)| | 👋 Gesture Detection | +------------------------------+ | Input events ("Help" / Reminder / Query) v +----------------------------------+ | 🗺️ Rules Engine: | | 🚨 Emergency alert | | ⏰ Read reminder | | ❓ Info query | +----------------------------------+ | Output v +-----------------------------+ | 🗣️ Text-to-Speech (TTS) | | 🔄 Fallback: ElevenLabs API| +-----------------------------+

text

+-----------------------------+

🦾 EcareBots: Accessible App [🎤 Mic] Speak your request [📹 Gesture] Wave for help [🕑 Reminders] "Take medicine @ 5pm" "Appointment: Tomorrow" +-----------------------------+ [Status: Listening...] [Output: 'Alert sent!'] +-----------------------------+ text

• 🗣️ Voice input: Real-time STT via Web Speech API.
• 👋 Gesture input: Wave/motion detection (upgrade path: MediaPipe/TFJS handpose).
• 🗣️ Spoken output: TTS (browser or ElevenLabs).
• 🗺️ Rules engine: Maps voice/gesture triggers to useful actions (alert, time, confirmation).

• 🚨 Emergency alerts: Say/gesture "help" to notify caregivers.
• 🕑 Reminders: Spoken med or appointment prompts.
• ❓ Information queries: "What's the time?", "Schedule?".
• 🏠 Accessibility-first UI: Fits assisted-living, home care, or mobility aids.

1️⃣ Install dependencies npm install

2️⃣ Run locally npm run dev

3️⃣ Open site in browser, allow mic/cam access text

Production deploy:

• Vercel: Build command npm run build, Output: dist/
• 🌐 No API keys in source! Use env vars for CI/CD secrets.

• Browser-first: All streams, STT, gestures run locally for demo/MVP.
• Serverless backend: Proxy for ElevenLabs TTS/STT for security/high-accuracy.
• Motion heuristic: Quick demo, replaceable w/ MediaPipe Hands/TFJS for full handpose.
• Pluggable models: Upgrade to Whisper, ElevenLabs STT/TTS as needed.

Short-term (1–4 weeks):

• 🤚 MediaPipe/TFJS gestures
• 🗣️ Whisper/ElevenLabs STT proxy
• ✅ E2E tests, deeper ARIA/accessibility
• 🔒 Secure ElevenLabs TTS integration

Medium-term (1–3 months):

• 👨‍👩‍👧 Multi-user/caregiver dashboard
• ⚙️ Configurable rules engine
• 📱 Mobile, PWA, offline mode

Long-term (3–9 months):

• 📞 Medical alert / SMS / telephony
• 🏠 Sensor integrations (ambient, motion)
• 🔏 Privacy/audit enhancements

Built for No Keyboards Allowed Hackathon (info).
Focus:

• 👓 Accessibility for care
• 💡 Demo meaningful hands-free interactions
• 🔌 Easy extensibility for AI/ML upgrades

• 🤔 Open issues for ideas/bugs/features
• 🚀 PRs to main w/ clear description, tests

MIT — see LICENSE

• Made with Vite, React, browser Web APIs
• Demo-ready: Web Speech API + webcam
• Inspired by hackathon mentors, accessibility advocates

• Issues/support: GitHub Issues
• Deployment: Live demo