Demonstration Case Study
Profile: 34-year-old marketing coordinator, remote work, relationship of 3 years, no children, lives in mid-sized city.
Presenting Pattern: Alex notices recurring Sunday evening tension and has mentioned "walking on eggshells" during video calls with partner.
"Okay, it's Sunday, 7:30 PM. Tomorrow's workday. Partner just asked what I want for dinner and I felt... annoyed? Which doesn't make sense. They were being nice. But I snapped a little, said 'I don't care, whatever.' And now they're quiet in the other room. This happens almost every Sunday. I don't know why. I'm tired but I slept fine. Work week coming, I guess. But it feels bigger than that. Like... dread? About nothing specific. Just... heavy."
Observation: The "Sunday evening" recurrence is significant.
Pattern: Pre-workweek anticipation coupled with relationship friction.
What this suggests: The "dread" may not be about the partner specifically, but the Sunday→Monday transition activates a stress response that spills into interactions.
Question worth asking: Is the heaviness about work, or about a feeling that the weekend "should have been" different?
Observation: Alex used this phrase in a previous memo about work calls, now applies it to home.
Pattern: Hyper-vigilance language transferring across contexts.
What this suggests: The nervous system may be maintaining a heightened state across environments, not just one.
Question worth asking: Where did Alex first learn to monitor others' moods closely?
Observation: "They were being nice. But I snapped."
Pattern: Emotional response mismatched to trigger intensity.
What this suggests: The Sunday trigger may be accessing something deeper than the present moment — a stored pattern, not a current event.
Question worth asking: What did Sunday evenings mean in Alex's earlier life?
System: Raspberry Pi 5, 8GB RAM — running continuously on a desk
During transcription: CPU at 60-80%, fans audible, temperature monitored. One-minute audio processes in ~30 seconds.
For files >2 minutes: Automatic chunking into 90-second segments. Prevents memory pressure (system has 7.9GB total, 4.2GB available at idle).
The cloud part: Pattern recognition happens locally. Analysis/narrative requires cloud augmentation — Pi cannot run large language model alongside transcription and web services.
The trade-off: Slower than cloud-only services. Audio never leaves your device. Pattern recognition happens in isolation. Understanding requires connection. This is the honest limitation.
Alex sends voice memo via secure channel. Audio transcribed using on-device processing (no cloud storage of raw audio).
System analyzes for:
GrowthXYNZ framework applied:
Analysis returned to Alex via preferred channel. No clinical language. No diagnosis. Just patterns Alex mentioned, organized in a way that might reveal what Alex already knows but hasn't seen clearly.
Not diagnosis: Alex is not "disordered." Alex has a pattern that causes friction. Patterns can shift.
Not therapy: AETHER organizes observations. It does not treat, heal, or replace professional care.
Not prediction: The framework suggests possibilities, not certainties.
Data handling: Voice memos transcribed locally, stored encrypted, analyzed in isolated environment.
Retention: User controls. Can delete anytime. Default: retain for pattern recognition across time.
Human review: No. This is automated pattern recognition, not clinical assessment.
Hand-off ready: If Alex ever wants to share this analysis with a counselor or therapist, structured format supports that conversation.
Alex's Sunday pattern is common. The dread-before-Monday, the relationship friction, the self-blame — these are human experiences, not disorders.
AETHER simply notices what Alex might miss while living inside the week.
The system sees the pattern. Alex decides what to do with it.