MeetCompanion — AI-Enabled Social Support Platform for Older Adults

1. Overview

Led a graduate-level integrative design project to address social isolation among older Chinese adults by designing an AI-assisted, voice-first social participation platform. Responsible for end-to-end product thinking from problem framing, concept design, to interactive UI prototyping using Figma AI in a 3-person team.

2. STAR Breakdown

Situation

Social isolation among older Chinese adults is a large-scale and growing societal issue, driven by geographic separation from children, low digital literacy, and overly complex mainstream social apps like WeChat. Existing tools fail to balance usability, emotional relevance, and autonomy, leading to low adoption and continued exclusion from community activities that could improve mental and physical health.

Task

Design a digital product that enables older adults to:

• Access socially meaningful local activities
• Interact with minimal cognitive burden
• Maintain independence while optionally receiving family support

The challenge was to translate behavioral theory (TAM, SST, Scaffolding Theory) into a concrete, usable product design under constraints of limited senior digital literacy and mixed-language presentation requirements for academic evaluation.

Action

• Conducted structured problem decomposition using TAM (ease of use), SST (emotional motivation), and Scaffolding Theory (learning progression) to define three core product principles.
• Defined product strategy around voice-first interaction, emotionally meaningful local activities, and adaptive scaffolding that fades over time.
• Designed end-to-end user flows including:
  • Daily AI-generated voice briefings for activity discovery
  • One-tap or preset AI-generated responses for low-friction communication
  • Family assistance mode where children can support discovery but cannot override elder decisions
• Built interactive UI prototype in Figma AI with a focus on reducing cognitive load and simplifying navigation architecture into four primary modules (Briefing, Activities, Messages, Settings).
• Iterated design based on usability issues:
  • Reduced visual strain by replacing high-saturation colors with a calm deep-blue palette
  • Removed bilingual clutter by introducing explicit language selection at entry point
  • Added handwriting input as fallback to support non-voice accessibility scenarios and edge cases where voice input fails or is uncomfortable
• Balanced competing constraints between simplicity (for seniors), expressiveness (for communication), and autonomy (to avoid family over-dependence).

Result

Delivered a fully interactive prototype and product specification that demonstrates a scalable design framework for elder-friendly digital participation tools. The final design integrates AI-driven interaction, family-assisted scaffolding, and low-friction activity discovery into a unified system aligned with behavioral and cognitive aging theories.

3. Product Thinking Highlights

• Identified that the core barrier was not motivation but cognitive and interaction complexity in existing social platforms
• Translated behavioral theories (TAM / SST / Scaffolding Theory) directly into product principles and UI decisions
• Prioritized voice-first interaction as a primary input modality, not a secondary accessibility feature
• Designed family involvement as a bounded support system, preventing over-dependence while enabling onboarding
• Used progressive disclosure (preset replies → free voice input) to implement skill scaffolding over time
• Reduced cognitive load through architecture simplification and elimination of redundant UI elements (e.g., bilingual duplication)

4. Challenges & How I Solved Them

• High cognitive load from initial UI complexity

  • Initial prototype included multi-language display and dense interface elements that overwhelmed users
  • Resolved by introducing single-language selection at entry and reducing navigation depth to a shallow 4-module structure

• Accessibility limitations of voice-only interaction

  • Voice input alone was insufficient for edge cases such as accent variability or contextual inappropriateness
  • Addressed by adding handwriting input as a secondary modality grounded in familiar motor memory patterns

• Balancing autonomy vs. family assistance

  • Risk of family mode turning into full substitution of elder decision-making
  • Designed constraints so family members can suggest and monitor activity participation but cannot act on behalf of users

• Design trade-off between academic completeness and usability clarity

  • Required inclusion of theoretical framing for academic evaluation while maintaining practical usability
  • Solved by embedding theory directly into product principles rather than treating it as separate documentation

5. Impact Summary

• Produced a complete end-to-end product concept from research → principles → UX flows → interactive prototype
• Established a scalable design framework for elderly digital inclusion products
• Translated multi-theory behavioral research into actionable product features
• Delivered a simplified interaction model optimized for low digital literacy users
• Demonstrated system-level thinking across UX, AI interaction design, and human support systems

6. Reflection

This project reinforced that for AI-driven or socially impactful products, the primary challenge is rarely feature design but alignment between human cognition, motivation, and interaction cost. The most important product decisions came from removing complexity rather than adding functionality. It also highlighted how AI should be positioned not as a “feature layer,” but as a scaffolding mechanism that gradually transfers capability to users while preserving autonomy.