WalkieTalkie
Nov 25
Background
User Research
Survey responses indicate that most participants lack a structured approach to self-reflection. Follow-up probing further revealed that they tend to respond in an ad hoc and reactive manner—often only when stress accumulates or specific problems become salient—and, instead of working through these experiences, they frequently cope by diverting their attention, for example by scrolling on their phones.
Participants also experimented with other AI tools for reflection. Although these interactions sometimes helped them make short-term progress, the tools were difficult to sustain as part of a long-term practice, and many of the apps bundled additional functions or features that diluted the reflective experience.
User Persona
User Characteristics: High cognitive and emotional load, comfortable with AI companion
Context of Use: Micro-transitions & everyday walking, emotional + reflective contexts
Problem Statement
1. Can automatic summarization (keywords, stories, trails) increase users’ awareness of emotional patterns?
2. How does state-aware AI dialogue affect the depth of user reflection during casual walking?
3. How does incorporating contextual signals (e.g., GPS trajectory, environment type, and user–character walk snapshots) affect the relevance of AI responses and users’ engagement with the reflection experience?
Existing workarounds:Tolan, PlanCoach, 林间疗愈室
Why They're Inadequate:Current solutions seldom retain and distill users’ reflections into a cumulative self-narrative and remain largely screen- and text-based, rarely engaging the body or context (e.g., walking)
User Journey
Technical Pipeline
App Key Features
Low-Fidelity Prototype
High- Fidelity Prototype