Health Companion
A health-tracking concept designed around state understanding, abnormal-signal awareness and calm, contextual guidance.
Challenge
What needed to be solved
Health information is often fragmented across symptoms, routines, emotions and clinical follow-ups. A useful product needs to preserve context without pretending to replace professional care.
Approach
How the product was framed
The product is structured around a repeating state loop: capture context, assess current status, identify abnormal signals, surface practical next steps and retain the history for future comparison.
01 · Product logic
Start with state, not content
The experience begins by helping the user describe what is happening now. The system then connects symptoms, duration, routines and recent changes before presenting guidance.
- Structured daily check-ins
- Abnormal-signal detection with explicit limits
- Context carried across multiple days
- Clear escalation guidance when risk is higher
02 · Experience
Separate assessment from support
The interface distinguishes between understanding the current condition, monitoring abnormal stages, and offering lifestyle or emotional support. This keeps important signals from being buried under general wellness content.
03 · AI role
Use AI to organise context
AI is used to summarise patterns, ask the next useful question and generate a concise status view. Rules and safety boundaries remain visible so the product does not imply medical certainty.
Outcome
The concept produced a clearer MVP boundary: structured input, state assessment, context history and safe next-step guidance before adding broader lifestyle features.
Learning
In health products, trust comes from explicit boundaries and continuity. The most valuable AI behaviour is often not answering more questions, but organising the situation well enough for the user to decide what to do next.