Experience Sleep Architecture Local AI Coach Zen & Audio Zero Cloud Architecture Privacy Policy Terms of Service Author (Novartus) ↗ GitHub Repository ↗

Download v1.0 APK

Engineered to know
your living body.

Autonomous personal health intelligence unifying offline Android Health Connect biometrics, 4-stage sleep polysomnography, nocturnal pulse dip, and local Gemini Nano coaching.

Cloud Leakage · 100% Local

4-Stage

Clinical Polysomnography

14.2%

Nocturnal HR Dip Tracking

<18ms

On-Device NPU Inference

Download v1.0 APK Explore Live Interface

OdinEye Titanium Android App & Smart Ring AIR

Nocturnal Dip: 14.2% Parasympathetic Equilibrium

Android AICore · Gemini Nano Offline Neural Voice Coach

AES-256 Encrypted Vault Zero External Servers · Mode 0700

Directly Integrated with Open Android Standards

Health Connect

Wear OS Telemetry

Smart Ring AIR

Android AICore

MIT Open Source

Interactive App Showcase

Ten unedited screens. One unified ecosystem.

Explore production captures directly from the running OdinEye Android build. Select any feature domain below to inspect the interface.

DOMAIN 01 · SYSTEM DASHBOARD

Real-Time Daily Equilibrium & Adherence

The OdinEye home dashboard synthesizes four primary biometric indicators into a single unified recovery score. View active cardiovascular exertion, daily medication adherence checklists, and quick-launch zen breathing triggers without opening secondary tabs.

✦

Restorative Recovery Gauge: Dynamic SVG radial score factoring sleep architecture and nocturnal autonomic dip.

✦

1-Tap Medication Check: Native offline adherence tracking synchronized with system notification alarms.

✦

Zero Latency Health Connect Sync: Instant telemetry ingest from Pixel Watch, Galaxy Watch, and Ring AIR.

Viewing Screen:

Full Production Gallery Click any capture to inspect in the flagship stage above

Home Snapshot

Action Hub

Biometric Rings

Sleep & Nocturnal HR

Gemini Nano AI

Tactile Breathwork

Reactive Mood

Health Targets

Source Privacy

Encrypted Vault

Clinical Sleep Science

Full 4-stage architecture.
Correlated with nocturnal pulse.

Standard sleep trackers lump your night into generic sleep and wake periods. OdinEye processes native sleep stage records from Health Connect into clinically validated Deep (N3), REM, Light (N1+N2), and Awake intervals, while computing your true parasympathetic heart rate dip.

Full 4-Stage Architecture 7h 42m Total · 94% Efficiency

11:30 PM (Sleep Onset) 03:15 AM (Peak Deep Slow-Wave) 07:12 AM (Wake)

Deep Sleep (N3)

1h 42m 22%

Cellular repair, growth hormone release & glymphatic neuro-waste clearance.

REM Sleep

1h 51m 24%

Procedural memory synthesis, motor-skill integration & emotional recalibration.

Light Sleep

3h 41m 48%

Circadian thermoregulation (-0.8°C core dip) & foundational metabolic reset.

Micro-Arousals

28m 6%

4 brief unconscious posture adjustments. Normal normative sleep fragmentation.

OdinEye Clinical Polysomnography Screenshot

Production Capture: Vitals Screen 4-Stage Architecture breakdown & Nocturnal Pulse Matching

OdinEye Dedicated On-Device AI Health Coach Screenshot

Production Capture: Dedicated AI Coach Offline SpeechRecognizer + Android AICore Gemini Nano

100% OFFLINE NPU INFERENCE · GEMINI NANO

Hands-free voice coaching.
Zero cloud leakage.

Ask complex clinical questions while stretching or cooking. OdinEye features a custom Kotlin OdinEyeVoiceModule for hands-free offline speech transcription and local text-to-speech, routed directly to Android AICore.

Test Simulated Gemini Nano Clinical Query: <18ms NPU Latency

⚡ OdinEye Health Engine (Gemini Nano) LOCAL NPU

"Your 88% recovery score is supported by 1h 42m of slow-wave Deep Sleep (22%) and a healthy 14.2% nocturnal cardiovascular dip down to 48 bpm. Your autonomic nervous system is in optimal parasympathetic equilibrium."

No Cloud Tokens Never calls OpenAI, Claude, or remote servers.

Private Biometrics Your heart rate and sleep are never uploaded or mined.

Instant Availability Functions completely offline in airplane mode or subways.

Acoustic Neuroscience

Mathematical frequencies.
No compressed audio loops.

Traditional wellness apps stream repetitive, compressed MP3 recordings. OdinEye synthesizes pure mathematical frequencies directly in memory using native 44.1 kHz PCM stereo buffers, coupled with rhythmic somatic haptics for eyes-closed breathwork.

Audio continues playing seamlessly in the background when your phone screen turns off.

Production Captures: Zen Sanctuary Tactile Breathwork Chamber & Reactive Dynamic Mood Face

Radical Privacy Architecture

Your biological telemetry belongs exclusively to you.

No cloud accounts. No remote databases. No commercial telemetry brokering. Here is how OdinEye compares to commercial wearable platforms.

← Swipe table horizontally to compare specifications →

Privacy & System Metric	OdinEye Architecture	Commercial Wearable Apps
Biometric Data Location Where your pulse, sleep, and heart rates are stored	100% On-Device Encrypted Vault Hardware Keystore + AES-256	Remote Cloud Datacenter Centralized multi-tenant databases
AI Model Inference How your recovery score and summaries are computed	Local NPU via Android AICore Gemini Nano runs strictly on-device	Third-Party Cloud APIs Telemetry sent across public internet
Health Connect Synchronization Accessing data from wearables and smart rings	Read-Only Offline IPC No telemetry ever leaves your phone	Cloud OAuth Upload Syncs your data to company cloud servers
Monetization & Advertising Tracking cookies, ad networks, and behavioral profiling	Zero Tracking · Zero Ads No analytics SDKs or telemetry miners	Commercial Telemetry Brokerage Aggregated user profiling & analytics
Licensing & Freedom Source code auditability and independence	100% Free & Open Source Permissive MIT License	Proprietary Subscription Paid paywalls & vendor lock-in

Engineering Specifications

Built to modern Android enterprise standards.

Developed in strict TypeScript 6 with 0 compiler warnings, native Kotlin modules, and hardware-level encryption.

Runtime & Stack

React Native: v0.86.3 modern architecture
Expo SDK: v57 enterprise framework
TypeScript: v6.0 strict mode (0 errors, 0 unused vars)
Target OS: Android 10+ (API Level 29 through 35)
Build System: Gradle 8.x with Hermes V8 engine

Native Android Modules

Voice Pipeline: Kotlin OdinEyeVoiceModule
Audio Engine: Kotlin OdinAudioEngine (PCM 44.1kHz)
Home Widgets: Native RemoteViews (4×2 layout)
Push Alarms: Android AlarmManager exact timers
Health Telemetry: Android Health Connect API 4.1.3

Cryptographic Isolation

Storage Engine: Hardware Keystore backed storage
Cipher Algorithm: AES-256-CBC with PBKDF2 (10k rounds)
Message Integrity: HMAC-SHA256 signature verification
File Permissions: Linux Mode 0700 private sandbox
Network Policy: Strict cleartext traffic prohibited

Release v1.0.0 Ready

Take back control of your health intelligence.

Experience personal health telemetry without compromising your personal privacy. Download the official signed APK or compile from source.

Download OdinEye v1.0 APK Inspect GitHub Repository

✔ Signed SHA-256 Checksum ✔ Free & Open Source (MIT) ✔ Zero Analytics SDKs ✔ Android 10+ Compatible

Engineered to know your living body.