Reflective Wellness · ~2025

AuraLLM

Know Your Energy. Protect Your Light.

Voice patterns. Body signals. The triggers you choose. Woven into a living portrait of your wellness — private, on-device, and never medical. AuraLLM combines voice energy analysis, habit tracking, and self-reported triggers into a single energy portrait powered by on-device ML, a dual-ring aura dial, chakra-mapped guidance, and an open AR glasses schema.

3
Signal Sources
0
Raw Audio Stored
6
Chakra Levels
100%
Offline Capable
AuraLLM energy dial interface showing dual-ring aura visualization and chakra guidance

Investor Summary

AuraLLM occupies a unique position in the $7B+ wellness app market by combining voice-derived energy analysis, structured habit tracking, and an open AR glasses schema into a single privacy-first experience. Unlike fitness trackers that focus on physical metrics or meditation apps limited to guided sessions, AuraLLM reflects how users feel and how they show up through a dual-ring energy dial that separates internal state from outward impact.

The core pipeline runs entirely on-device. Voice samples are processed in-memory and immediately discarded — only numeric scores persist. Habit entries stay in the local Room database. The 3-day rolling analysis window runs via WorkManager every six hours, generating reflective insights through a pluggable LLM interface that currently uses a deterministic template engine with a clear roadmap to on-device TFLite and cloud providers. This architecture means the product delivers full value at zero marginal cost per user while the freemium model ($4.99/month Aura+) gates advanced analytics, custom analysis windows, and AR visualization without restricting core privacy guarantees.

The AR glasses integration — built around an open JSON schema for DialState — positions AuraLLM ahead of the curve as smart glasses move from niche hardware into mainstream consumer adoption. The pulsing aura ring, energy score, and chakra bar render on any compatible HUD, creating a persistent wellness overlay that no competing app currently offers. Early integration with this form factor builds brand association and switching costs before the market matures.

Person meditating with AR glasses showing colorful aura energy rings

Product Capabilities

  • ✓ Voice energy extraction (speaking rate, tone warmth, vocal energy) processed entirely on-device with zero raw audio persistence
  • ✓ Dual-ring aura dial separating Self Energy from Others Impact with colour-coded labels (Radiant, Steady, Recharging, Red Zone)
  • ✓ Six-category habit tracking with quality ratings and a transparent Habit Bonus algorithm influencing energy scores
  • ✓ Trigger tags — Sleep, Exercise, Social, Work, Nature, Creative — surfacing connections you feel but can't name
  • ✓ Chakra guidance system mapping energy state to Root-through-Crown levels with three actionable micro-interventions per level
  • ✓ 3-day rolling trend analysis via WorkManager with pluggable LLM insight generation (template, TFLite, GGUF, cloud roadmap)
  • ✓ Red Zone crisis self-report — never algorithmically inferred — with grounding-focused guidance and trigger logging
  • ✓ Open DialState JSON schema for AR glasses HUD with BLE broadcast and companion SDK planned

Three Signals. One Story.

No other wellness app combines this depth of personal energy reflection while keeping every byte on your device. AuraLLM weaves voice analysis, habit tracking, and self-reported triggers into a single, living energy portrait.

AuraLLM habit tracking and voice analysis interfaces

Voice Energy — Your Tone Tells a Story

With explicit, revocable consent, AuraLLM captures a brief voice sample and extracts derived features entirely on-device. Raw audio is processed in-memory and immediately discarded — only numeric scores persist: Speaking Rate (words-per-minute cadence), Tone Warmth (emotional warmth 0-100), and Vocal Energy (estimated energy 0-100). Future TFLite prosody model integration will deepen analysis without changing the privacy guarantees.

Habit Tracking — Small Actions, Big Impact

Log the habits that shape your energy every day across six categories: Meal, Sleep, Exercise, Hydration, Meditation, and Social. Each entry is rated 1-5 for quality. A rolling 3-day average drives the Habit Bonus: excellent habits (4-5) add +10 to +15 points to Others Impact, good habits (3-4) add +5 to +10, building habits (1-3) add +0 to +5. Consistent positive actions compound visibly over days.

Trigger Tags — Name What Moves You

Tag what's influencing your energy today and let the model surface connections you feel but can't name. Six tag categories — Sleep, Exercise, Social, Work, Nature, and Creative — capture the dimensions of daily life that shape your energy portrait. The LLM weaves these triggers into personalized reflective narratives, revealing patterns across days and weeks that manual journaling would miss.

The Dual-Ring Energy Dial

At the heart of AuraLLM is a concentric visualization that separates how you feel from how you show up. Two rings, one aura score — a reflective lens, not a clinical verdict.

AuraLLM dual-ring energy dial showing Self Energy and Others Impact rings

Two Dimensions of Energy

The outer ring captures Self Energy — how you feel internally, derived from self-report or voice analysis. The inner ring captures Others Impact — how you show up to the world, calculated from self-energy plus a habit bonus that rewards consistent positive actions.

Aura Labels

Radiant 70-100
Steady 45-69
Recharging 20-44
Red Zone 0-19

The Aura Colour Spectrum

Every energy level maps to a DMT-palette colour that pulses through the dial and glasses HUD:

80-100 Green 60-79 Purple 40-59 Cyan 20-39 Yellow 0-19 Red

Chakra Guidance & Red Zone

Energy maps to a structured framework for micro-interventions. Each chakra level delivers three specific, actionable suggestions. The Red Zone is a conscious user action — never algorithmically inferred.

Wellness coach analyzing energy readings on a futuristic tablet
☯

Crown (85+)

Purpose & stillness. Sit in stillness for 3-5 minutes. Offer a kind act. Name one purpose to embody today.

☯

Throat (70-84)

Expression & communication. Speak one honest sentence kindly. Hydrate and take slow breaths.

♥

Heart (55-69)

Gratitude & connection. Place a hand on your chest and breathe in 4, out 6. List three things you're grateful for.

☼

Solar (40-54)

Clarity & action. Identify one thing you can control right now. Do it with intention.

♬

Sacral (25-39)

Creativity & movement. Move your body for 5 minutes. Create something small. Connect with water.

⚖

Root (0-24)

Grounding & survival. Take a slow grounding walk. Exhale longer than you inhale. Choose one steadying habit.

Red Zone — Crisis Self-Report

The Red Zone is never algorithmically inferred. AuraLLM does not claim to detect distress, diagnose conditions, or assess risk. Red Zone activation is a conscious user action — an honest self-report that says "I'm not okay right now." When activated: energy scores set to minimum (5/5), aura label shifts to Red Zone, chakra drops to Root with grounding suggestions, the dial pulses red as a visible reminder, and the state is recorded with the trigger self-report:<reason>.

The app meets you where you are. It doesn't judge. It doesn't diagnose. It just reflects.

The chakra system also applies a self-centred penalty when Self Energy exceeds Others Impact by 20+ points and a red zone penalty — ensuring the system rewards outward awareness, not just personal highs.

Multi-Day Trend Analysis

A single snapshot tells you where you are. A rolling window tells you where you're headed. Background analysis runs every 6 hours via WorkManager, analyzing a 3-day rolling window of energy readings and habit data.

High Energy, Upward

"Your energy has been consistently high. Keep nurturing what fuels you."

Moderate, Rising

"Nice upward momentum — small positive changes are compounding."

Moderate, Falling

"Your energy has been tapering. A short reset activity might help."

High, Slight Dip

"Your energy is strong but shows a slight dip — consider what may have shifted."

Low, Improving

"Low overall but the direction is improving. Gentle consistency matters."

Low, Declining

"It looks like a challenging stretch. Be kind to yourself — rest is productive too."

All insights are appended with: (For wellness reflection only.)

LLM-Powered Reflective Insights

A pluggable LLM architecture generates personalized wellness narratives from your energy data. The LlmProvider interface accepts energy readings and habit summaries, returning reflective insights. Integration is completely optional — your energy calculations, habit tracking, and dial work perfectly without it.

Active

Mock (Template)

Deterministic, instant, zero-cost. No network required. Pattern-matched insights from energy data.

Planned

On-Device TFLite

Private, fast, no API key. Runs entirely on device for maximum privacy with richer prosody analysis.

Planned

On-Device GGUF

Maximum quality, fully local. Quantized language models running natively for deepest on-device narratives.

Planned

Cloud API

Deepest narratives, user-configured. Only aggregated scores sent — never raw audio or identifiable data.

LlmProvider Interface

interface LlmProvider {
    suspend fun generateInsight(
        readings: List<EnergyReading>,
        habitSummary: String? = null
    ): String?
}

Aura on Your Glasses

A minimal heads-up display shows your energy state at a glance — colour-coded, distraction-free, always private. Compatible with the open DialState schema for any smart glasses platform.

AuraLLM AR glasses HUD preview showing pulsing aura ring and chakra bar overlay

The Glasses HUD

  • ● Pulsing Aura Ring — Colour-coded concentric circles that breathe with a 2-second animation cycle
  • ● Energy Score — Combined average displayed as a large number
  • ● Aura Label — Radiant / Steady / Recharging / Red Zone
  • ● Chakra Bar — Vertical ladder showing current chakra level with active/inactive segments
  • ● Camera Passthrough — Live camera feed with HUD overlay for real-world AR preview

Transport Roadmap

Active
JSON Export
Planned
BLE Broadcast
Planned
Companion SDK
Planned
Shared File

Open DialState Schema (JSON)

{
    "selfEnergy": 78,
    "othersImpact": 68,
    "auraColorHex": "#9C1AE7",
    "label": "Radiant",
    "chakra": {
        "level": 4, "label": "Throat",
        "colorHex": "#3AE5E7",
        "suggestions": ["Speak one honest sentence kindly.",
            "Hydrate and take slow, open breaths.",
            "Write a short note of appreciation."]
    },
    "isRedZone": false,
    "timestampMillis": 1738800000000
}

A Visual Experience Rooted in Energy

AuraLLM features a DMT neon aesthetic on deep midnight — a design language inspired by bioluminescence, energy fields, and visual consciousness.

The DMT Palette

AuraViolet #9C1AE7
High energy, crown chakra
AuraTeal #3AE5E7
Clarity, throat chakra
AuraGreen #2CD27E
Positive energy, heart
AuraAmber #F5CF1D
Caution, solar/sacral
AuraRose #F7093B
Alert, root, red zone

Surfaces

Midnight #0F0B1A
Primary background
Card #1A1528
Elevated cards, panels
On Surface #F0ECF7
Primary text

Nine AI Illustrations

Created with FAL (gpt-image-1.5) at 1024x1024. Each illustration bridges DMT-inspired energy art with clean mobile interface design.

• Onboarding walkthrough
• Account creation
• Daily dashboard with dial
• Weekly trend chart
• Close-up aura dial
• Voice analysis results
• Habit logging interface
• Settings & privacy
• Marketing phone mockup

Typography: Cormorant (display serif) + Outfit (geometric sans)

Privacy Is the Product.

Your wellness data is deeply personal. AuraLLM was engineered from the ground up with privacy as a core architectural principle — not a feature, not an afterthought.

🔒
Local-First
Room database on-device. No cloud servers in the core data path.
🎤
No Raw Audio
Voice processed in-memory, immediately discarded. Only numeric scores stored.
☑
Consent-Gated
SpeechConsentManager via DataStore. Must be explicitly true before any capture. Revocable any time.
🚫
No Data Harvesting
No advertising profiles. No third-party data brokers. No analytics SDK.
📡
No Network (Core)
The entire core pipeline — scoring, dial, habits, speech — works fully offline.
⚠
Not Medical
All copy uses "reflective wellness insights." Red Zone is self-report only. No condition detection.
AuraLLM privacy-first architecture showing on-device processing and consent flow

Consent Flow

User taps "Enable Voice"
→ SpeechConsentManager.grant()
→ DataStore: consent = true
→ SpeechFeatureExtractor available

User taps "Revoke"
→ SpeechConsentManager.revoke()
→ DataStore: consent = false
→ Feature extraction disabled

Data Architecture

Clean Architecture with MVVM. Three core data entities in Room, a WorkManager-driven analysis pipeline, and a Compose UI layer connected through ViewModel state flows.

Data Flow Pipeline

User Input / Voice ──► SpeechFeatureExtractor ──► SpeechFeature (Room)
                                                        │
Manual Check-in ──► EnergyReading (Room) ◄──────────────┘
                          │
                    EnergyRepository (3-day rolling window)
                          │
                    AnalysisWorker (WorkManager, every 6h)
                          │
                    LlmProvider.generateInsight()
                          │
                    EnergyReading.insight (written back)
                          │
                    EnergyDialViewModel ──► UI (Compose)
                          │
                    DialState ──► AR Glasses (JSON export)

EnergyReading

Table: energy_readings

selfEnergy, othersImpact, triggers, insight, source, timestamp

SpeechFeature

Table: speech_features

speakingRate, toneWarmth, vocalEnergy, sampleDurationSec

HabitEntry

Table: habit_entries

category, label, quality (1-5), timestamp

Platform
Android 8.0+ (API 28)
UI
Jetpack Compose + Material 3
Camera
CameraX (core, camera2, lifecycle)
Build
Gradle (AGP 8.7.3), Kotlin

Platform Architecture

🎤

Voice Engine

On-Device Speech Analysis

SpeechFeatureExtractor processes audio in-memory with zero disk writes. Consent-gated via DataStore, the pipeline extracts speaking rate, tone warmth, and vocal energy. Raw audio discarded immediately. Future TFLite prosody model integration planned.

📈

Habit Engine

Structured Wellness Tracking

Room database stores HabitEntry records across six categories with 1-5 quality ratings. A 3-day rolling average drives the Habit Bonus algorithm that transparently influences Others Impact scores. Entries never leave the device.

🤖

LLM Integration

Pluggable Insight Generation

LlmProvider interface accepts energy readings and habit summaries. Currently backed by deterministic template engine (zero cost). Planned: on-device TFLite, GGUF, and optional cloud API. All insights append wellness reflection disclaimer.

👓

AR Glasses

Open DialState Schema

Serializable JSON object with selfEnergy, othersImpact, auraColorHex, label, chakra level and suggestions. In-app HUD preview with pulsing aura ring. Camera passthrough for AR preview. BLE broadcast and companion SDK planned.

🔒

Privacy-First

Architectural Guarantee

Local-first Room database. No raw audio on disk or network. SpeechConsentManager with explicit grant/revoke. No analytics SDK, no advertising profiles, no data brokers. Full offline capability. Red Zone is self-report only.

🕐

Background Analysis

WorkManager Scheduling

AnalysisWorker runs every six hours consuming a 3-day rolling window. Feeds aggregated scores to LlmProvider.generateInsight(), writes reflective insight back to the reading. Battery-aware scheduling respects Doze and App Standby.

Technology Stack

Kotlin Android SDK 28+ Jetpack Compose Material 3 Room Database WorkManager DataStore CameraX kotlinx-serialization Gradle (AGP 8.7.3) MVVM Clean Architecture TFLite (Planned) GGUF (Planned) BLE (Planned)

Differentiation and Moat

Privacy as Product

Not a feature — an architectural guarantee. Zero raw audio persistence, consent-gated features, no analytics SDK, no data brokers. In a market where wellness apps routinely harvest sensitive data, AuraLLM's privacy posture is a genuine competitive advantage.

AR-First Positioning

The open DialState JSON schema and glasses HUD preview position AuraLLM to ride the smart glasses adoption curve. Building the integration layer now establishes brand association and switching costs before mainstream hardware arrives.

Multi-Signal Synthesis

No competing app combines voice-derived energy, structured habit tracking, and self-reported triggers into a single composite score. The dual-ring dial creates a defensible product category rather than competing in an existing one.

Freemium with Premium Expansion

Revenue scales with user value, not data exploitation. Premium features enhance the experience without gating core privacy protections.

Free (Core)

$0
  • ✓ Voice energy analysis
  • ✓ Habit tracking (6 categories)
  • ✓ Daily energy dial
  • ✓ Trigger tags
  • ✓ 3-day rolling insights
  • ✓ Chakra guidance
  • ✓ Full local processing
  • ✓ Data export (JSON)

Premium (Aura+)

$4.99/mo
  • ✓ Everything in Free
  • ✓ Advanced trend analytics
  • ✓ Custom analysis windows
  • ✓ AR aura visualization
  • ✓ Multiple LLM insight styles
  • ✓ Priority feature access

Results and Impact

Production-grade mobile architecture with measurable engineering outcomes and a clear path from core product to premium monetization and AR hardware integration.

$0

Marginal cost per user

$4.99

Premium monthly revenue

100%

Offline capable

9

AI-generated illustrations

Freemium Unit Economics

Core pipeline runs at zero marginal cost — no cloud inference, no per-user storage, no API fees. Revenue scales with premium conversions, not data exploitation.

Not Medical — By Design

Deliberate positioning avoids regulatory complexity. "Reflective wellness insights" framing serves users who want structured self-reflection without clinical framing.

AR Hardware Ready

Open DialState schema positions for smart glasses adoption. Early integration builds brand association and switching costs before the AR wellness category matures.

Commercial Use Cases

Applicable across consumer wellness, corporate wellbeing programs, wearable hardware partnerships, and white-label licensing for health platforms.

Consumer Wellness

Direct-to-consumer Android app with freemium conversion to Aura+ subscription. Privacy-first positioning drives word-of-mouth in communities increasingly sceptical of data-harvesting wellness apps.

Corporate Wellbeing

White-label deployment for enterprise wellness programs. The local-first architecture means employers never touch personal wellness data — they provide the tool, employees own the data.

Wearable Partnerships

Smart glasses manufacturers get a ready-made HUD implementation. Ring and watch manufacturers can ingest DialState for ambient energy display. Early partnerships create distribution before the AR wellness category is crowded.

Ready to See Your Energy?

AuraLLM brings together voice analysis, habit tracking, self-reported triggers, a dual-ring energy dial, chakra guidance, and AR glasses integration into one privacy-first Android experience.

Know your energy. Protect your light.

Schedule a Demo View All Projects

For personal wellness reflection only. Not a medical or diagnostic tool.

In Action

Person meditating with AR glasses showing colorful aura energy rings
Wellness coach analyzing energy readings on a futuristic tablet
Runner with AR glasses showing real-time biometric aura overlay