Research & Roadmap
Behavioral audio awareness, privacy-first design, and clinically informed development.
Purpose
This page is intended for research partners, clinical collaborators, and grant reviewers.
SnapHabit LLC, the company behind AwareFlow™, is developing a system for detecting subtle behavioral signals and translating them into real-time awareness.
The work aligns with priorities in digital health, behavioral science, and human-computer interaction.
Important: AwareFlow is not a medical device and does not provide diagnosis or treatment. The research described here reflects areas of investigation, not clinical claims.
Problem & Opportunity
Subtle habits such as sniffing, throat clearing, or repetitive sounds are common and often unconscious.
These behaviors can impact focus, relationships, and perceived social tension — particularly in contexts such as misophonia.
- They often occur without awareness
- They may reflect nervous-system load or stress
- They can create disproportionate impact on others
There is currently no widely adopted tool that provides real-time awareness of these patterns without recording or storing audio.
Scientific Foundations
AwareFlow builds on established concepts across behavioral science and digital health.
-
Behavioral signal patterns
Short, repetitive actions can reflect internal physiological or emotional state.
-
Contextual modulation
Fatigue, environment, and emotional load influence frequency and intensity.
-
Digital phenotyping
Using timing, frequency, and pattern data instead of raw recordings to derive insight.
The central premise: these behaviors are not noise — they are measurable signals.
Research Aims
-
Safe pattern detection
Develop on-device models that identify short behavioral sounds without storing audio.
-
Context modeling
Explore relationships between behavior, time, environment, and subjective state.
-
Awareness-driven feedback
Design interventions that support noticing without inducing pressure or compliance behavior.
Approach
- On-device processing instead of cloud-based audio collection
- Pattern-level abstraction instead of raw data storage
- User-controlled feedback rather than automated correction systems
This approach reduces privacy risk while preserving meaningful behavioral insight.
Development Roadmap
-
Phase 1
Sniff-only detection with personalized calibration (current)
-
Phase 2
Expansion to additional behavioral sound classes
-
Phase 3
Context-aware insights incorporating environment and temporal patterns
-
Future work
Clinical partnerships and formal validation studies
Privacy & Ethics
The system is designed around strict privacy constraints.
- No raw audio storage
- No passive cloud collection
- Explicit, opt-in participation for any research sharing
The goal is to enable behavioral insight without compromising personal boundaries.
Collaboration
We are open to collaboration with researchers, institutions, and funding partners.
For inquiries or partnership discussions: